In a recent post, I said that we should treat ability as an output of our instruction, rather than an input, and went on to say that:

“Teaching only makes sense if we believe instruction changes what students know.”
- Caiti Wade

However, I am also a practicing classroom teacher, and it would be disingenuous of me to not acknowledge the practical challenge that exists here when we are somehow trying to match a single lesson to learners of a variety of different levels. As a mathematics teacher (a subject that is highly hierarchical) who has taught both grouped/streamed and mixed-prior-attainment classes, and as someone who works externally with a lot of schools/teachers, I am very aware of the fact that it’s easy to say we should teach for everyone, but much harder to manage practically. When it gets to a point where a student is 10 years into their schooling and they’re missing substantial prior knowledge, it raises some difficult decisions around what is actually achievable.

As part of managing this and picking up the pieces of previous instructional gaps, I’ve been working on my new instructional obsession:

Access questions!

I made that name up myself. The idea itself isn’t completely novel. We’ll get to that. We’ll also get to why I’ve packaged this concept under a new name, but call it whatever makes you happy. I’m by no means a marketing whiz, so I’m sure there are better names floating around the edu-world.

To briefly summarise an access question, it’s retrieval practice meets checking for understanding meets gapless instruction, and hopefully, avoids some dreaded lethal mutations of the above.

This post will combine elements of theory/research with practical implications. My rationale for this is of course two-fold:

1. If you know the underlying evidence-base, you can pair it with your professional expertise to apply it to different contexts/subjects;

2. Understanding the how, why and when to apply something is part of avoiding lethal mutations - that is, where the practice becomes largely performative, but doesn’t really meet the core instructional benefits that make it worthwhile.

I will give the caveat before I begin that at first this idea will possibly seem a little too simple, but I’d encourage you to engage a little deeper (if you read this, you will have little choice because I’ve really gone into detail).

the short version: an access question

Before teaching a piece of new content:

• Identify the prerequisite knowledge students need to access the new learning.

• Ask a short question that requires students to retrieve that knowledge immediately before they need it.

• Check the response of the whole class so you know whether students can actually access what comes next.

• Temper your prompts to keep retrieval effortful but successful where the knowledge is there but difficult to retrieve.

• Reteach when necessary & continue to check the responses of the class if the prerequisite knowledge genuinely isn’t there - until they get it!

• Move into the new learning only once you’ve established that students have enough of a foundation to access it.

Those questions that you create are access questions, but let’s avoid some lethal mutations and make sure we do them right by deep diving the theory!

make success possible for every student

Firstly, this strategy is inspired by a principle that underpins a lot of my presentations:

Make success possible.

What this really means is believing that ability is largely an output of instruction, rather than being a fixed input that students enter your classroom with. It’s the belief that all students are capable of learning and that our instruction holds the power to do this. Recently on a conference panel, I was asked for my ‘silver bullets’ for school improvement, and while I don’t believe that there is a silver bullet that can change a school on its own, the belief that we as educators can make success possible for our students through our instruction is probably the closest thing to a silver bullet we can have.

The challenge is that, as a practicing classroom teacher myself, I know that whilst this is a lovely sentiment, its practical realisation is hard to conceptualise.

For one, we see students enter our classrooms with different ‘abilities’ every day. Some students pick up on things quickly and others don’t, and some have diagnosed learning challenges that create variation in exactly how that learning process looks for them. We see how the same lesson lands completely differently for different students, and so it is a genuine challenge to see how it is that instruction can largely determine ability; however, this is not a new idea that I’ve self-generated in hopes it will come across morally profound. For instance, Siegfried Engelmann said:

Or, Graham Nuthall in Hidden Lives of Learners:

Or, more recently, David Didau (who will feature here a lot because his work inspired a good portion of this):

In essence though, this isn’t about there not being differences in individual learners, but rather viewing these differences as something that we can address through our instruction. As Didau points out, this is a useful mindset to have when teaching, even if it isn’t strictly true all the time (although he words this far more eloquently than I have here, before we take this out of context!). There are of course several factors that can impact upon a student’s ability to access or experience success with the curriculum, including genetic factors, absences or even behaviour; however, it’s far more useful for our focus to remain on the one thing we can control as teachers. Classroom instruction.

I can make this a little simpler though:

Knowledge builds on knowledge, so let’s build up some knowledge!

Since we know that new knowledge needs to connect to existing knowledge held in long-term memory (aka prior knowledge), building up this network of prior knowledge is a pretty powerful thing we can do for our students. You’ll have likely experienced that feeling where it’s easier to learn something new about a topic you already have a lot of knowledge about, than it is to learn about something you don’t know anything about. As such, building up this bank of prior knowledge is going to support future learning for our students.

To put this into a nice diagram, below I have a web of golden ‘dots’. This represents prior knowledge. This knowledge is stored in ‘schemas’, which are essentially connected webs or mindmaps of information. When new knowledge enters our working memory, we need to contextualise it within what we already know - something that is far easier if we know more stuff!

Now add to this the fact some knowledge is hierarchical, like knowing how to write a sentence must precede knowing how to write a paragraph. If students are missing the prerequisite prior knowledge, learning something new is going to pretty hard - or even impossible.

In summary, having the right prior knowledge both makes new learning accessible and makes new learning easier.

But how do we get our students to have the right prior knowledge?

Didau proposes a practical solution for this that he terms gapless instruction, which can be summarised as finding gaps in our teaching & filling them in. Didau’s post will go into far more detail than mine will here, but he emphasises that the core principles of gapless instruction are as follows (paraphrased from this post):

1. Assume that student failure is the result of gaps in instruction

2. Success needs to precede struggle

3. Aim to make sure the practice students engage in is ‘perfect’ (ie. they’re not practicing the wrong stuff!)

What I’m going to do today is draw upon these principles + a couple of others to tell you about Access Questions, which is part of me actively working towards gapless instruction.

access questions: the core principles

An access question is a ‘check for understanding’ question that comes immediately before any core content in the lesson, used to check whether students have the prerequisite knowledge needed to access the new learning. Think gapless instruction + retrieval practice.

That sounds simple, but there’s quite a few mechanisms at play with Access Questions that make them effective - and prevent lethal mutations.

First, this idea is underpinned by two principles that form the basis for most of my lessons:

1. Make sure your students are thinking;

2. Make sure they’re thinking the right stuff.

Make sure your students are thinking.

This one seems obvious. Firstly, students need to be cognitively engaged with whatever learning they’re taking part in, but, secondly, that learning should be at an appropriate level of difficulty for them. I refer to this as the ‘Goldilocks Zone’, but it draws on a few different areas of research.

Firstly, in retrieval practice, an effective retrieval should be effortful and successful to be a desirable difficulty. Retrievals that require little effort are remarkably inefficient at supporting learning, compared to those that are challenging, but still successful1. Secondly, research indicates that we are also not particularly good at judging when we’re learning2. That is, we tend to perceive that learning should be easy, and when it is, we mistakenly believe that we are learning more than when some challenge is involved. This is why students often feel they learn a lot with private tutors, but don’t retain the information come assessment time - the sessions will usually feel easy in the moment, but they don’t lead to durable learning.

The first principle is around ensuring that students are hitting this appropriate zone of mental effort: an area where learning isn’t too difficult that they’re not able to experience success, but not so easy that it’s inefficient or ineffective.

Obviously, we want to make sure students are putting in appropriate amounts of cognitive effort throughout our lessons.

Make sure they’re thinking the right stuff.

The next principle is make sure that students are thinking the right stuff. As Hollingsworth and Ybarra say, “practice makes permanent”. If students practice the wrong stuff, they learn the wrong stuff. This really just builds the case for frequent short-cycle formative assessment to monitor student practice closely and check that they’re understanding the learning correctly.

There are several techniques for frequent short-cycle formative assessment, such as mini whiteboards, picking a non-volunteer, choral response or others (see here for my crazy detailed post on CFU, or here for Shaun Brien’s strat chat on CFU). To really emphasise how much CFU you can use (keeping in mind content differences), I’ve previously spoken about how my students would spend around 90% of their lessons on mini whiteboards, and the below is a screenshot of my slide desk, showing over 100 slides for 70 minutes of lesson time.

To be clear, I don’t expect to necessarily go through every single one of the 100+ slides above - they exist in case I need them to make sure my students are thinking the right stuff.

the gapless instruction piece

Using the two principles above, access questions combine components of gapless instruction with effective retrieval practice. Let’s focus on gapless instruction first, and let’s see some examples - not just maths! Before I give you examples, I will preface this by saying that the examples are there to give us something to refer to at this stage in the post - they don’t tell the whole story, and at this stage, will likely seem trivial and silly. Bear with me.

1) Grant access

Let’s say the main learning for a particular lesson is around completing a written task based on the following prompt:

The first piece in access questions is around identifying prerequisite knowledge that students require to even access that activity in the first place. This is about identifying the barriers that students might have to accessing the core learning for the lesson. There are of course lots of things that might prevent students from being able to engage in the new teaching above (writing sentences/paragraphs, for instance!), so we’re going to be selective in what potential barriers we identify. For this content, I’m going to assume my students can write - largely because if they can’t, that’s not something I can just casually slot in before this. However, I am going to assume they need some knowledge about the Titanic. Immediately before teaching this within my lesson, I’m going to slot in the following question:

This is a pretty basic retrieval question (or access question), asking students to retrieve the basic facts required for the core learning above. If my students don’t know any of these pieces of information, that’s a good opportunity for me to pause my lesson there and reteach this content.

However, I likely wouldn’t stop there. The next access question I would ask would get students to start to connect the relationships between the facts above, in such a way that it supports building up a case for the core teaching of the lesson:

The above expands on the initial retrieval of facts by getting students to connect these facts. I’ve used sentence frames to ensure that students fully answer the question, because my focus here is very much making sure students understand how these events combined to contribute to the loss of life on the Titanic.

Again, if my students didn’t know this, I would stop and use this as a teaching opportunity. The whole purpose of this section is find the barriers - likely from previous instruction - that would prevent students from accessing the lesson content, and teaching them if you need to.

2) Activate prior knowledge

One of the core benefits of this technique is that it activates prior knowledge.

“Activating some relevant prior knowledge before instruction is beneficial for students of all levels of prior knowledge.”
- Craig Barton

If we refer to my image from above, new knowledge needs to have a place to slot into within existing prior knowledge stored in long-term memory.

This technique is entirely about connecting new knowledge to existing prior knowledge for the learner. Rather than a learner seeing new knowledge as being isolated, it makes explicit the connections between prior knowledge and new learner. It also prevents a novice learner from ‘guessing’ how new knowledge fits in with what they already know.

In essence, it’s about activating the relevant schemas to explicitly bridge the gap between existing knowledge and new knowledge.

Where an access question does this differently to some activating prior knowledge techniques I’ve seen is it ensures students are actively thinking about the material. By just telling our students where new knowledge fits in, we don’t necessarily achieve the level of cognitive engagement or effortful thinking needed to make sure they’re even processing what we’re saying. Remember, we need to make sure our students are thinking.

Of course, by running this as a check for understanding, we also make sure they’re thinking the right stuff. We want them to activate the right schemas, so new knowledge slots into existing knowledge correctly.

I have an embarrassing personal example of how I was actually slotting new knowledge into the wrong ‘place’. Up until somewhat recently, my knowledge of world history and geography was abysmal. A good friend of mine, James, had been teaching me some basic geography, starting with Italy, where one of the things he’d taught me was that Rome was the capital city (yep, that’s a real thing I didn’t know until I was an adult). One night at pub trivia3 , a question came up asking for the origin of something. We didn’t know the answer, but James suggested that it was likely Italian, given its links to the Roman Empire.

At this moment, I responded, asking, “But what does that have to do with Italy?”

The trivia team were confused. They had been witnesses to my mini-geography lessons, and knew that I knew that the capital of Italy was Rome. One of them replied pointing this out, “You know that the capital of Italy is Rome.”

I replied, “Wait, is Rome the city the same as Rome the country?”

You may be thinking, Caiti, there is no Rome the country. You’d be correct. What had happened was in my head, I essentially had two completely disconnected schemas for Rome, the capital of Italy, and the Roman Empire (which I had thought was a country - a fact I’m still not sure the origin of). To put this into a diagram:

Exactly how long I’d been building up these two totally separate cognitive schemas, I can’t tell you. What I can tell you is that my friends watching a couple of schemas collide and some misconceptions get broken apart live.

For novices, it’s important to make sure that we activate the right schema before they begin.

3) Manage cognitive load

The final piece in this is that idea of managing cognitive load in our teaching. When we teach new content to students, there’s often a lot of prior knowledge built in.

Take this question, for instance:

While my focus within this example would very much be on isolating ‘x’ using inverse operations applied to both sides, hopefully you can see that there’s actually a lot of prior knowledge required, but if not, see below:

I could have gone far more in-depth too with the prior knowledge embedded within the teaching of this question - I haven’t even referenced the place values within the number 19!!!

The reason I reference this is I’ve gotten into teaching examples, like the one above, in the past, and have felt as though I’m teaching several different things at once - because I am. Surely you know that feeling where you’re part way through teaching something and you realise they don’t know some necessary prior content, so suddenly you’re teaching content within content and it’s all a big mess??

Well, access questions exist to avoid that. By having students work with some of the relevant prior knowledge beforehand, it stops me needing to teach this knowledge within the middle of another example, but it also makes it obvious to the students what they already know. Instead of students wondering why 3 - 3 = 0, but 2/2 = 1, in the example above, they can instead focus their attention to the new learning: probably balancing the equation to isolate x.

Instead of getting halfway through the above example and having a student put their hand up to ask that question, I could build in some access questions before this to help them focus their efforts on the new content when I get to the main question:

The sequence of access questions above moves students through the core prerequisite knowledge that they will draw upon when I teach that example above. This means their focus can shift to really ‘zooming in on’ the steps required to actually solve an equation.

the retrieval practice piece

This is where I love the concept of an access question. Really, it’s just a very intentional application of retrieval practice. Whilst retrieval practice is by far one of the most effective things we can embed into our classroom practice, usually it’s done in such a way that it loses its effectiveness - a lethal mutation, one might say.

I am aware that doing the following runs the exact risk of over-simplifying the research around retrieval practice, but I summarised four key components of effective retrieval below:

Effortful

Recall that I said that one core goal of teaching is to ensure your students are thinking? What I really mean by that is effortful thinking.

Whilst this sort of retrieval is clearly part of what is commonly termed ‘activating prior knowledge’, the difference here is that by asking questions, we maintain some level of effort in the retrieval. Carl Hendrick explains this nicely:

“Retrieval practice works by strengthening memory traces that already exist. The act of pulling information out of long-term memory; of searching, effortfully reconstructing, generating an answer from partial cues, consolidates and stabilises what has been learned.”
- Carl Hendrick

The frustration that arises here is that there’s this counterintuitive idea in learning, captured nicely by ‘desirable difficulties’: learning that feels harder in the moment often leads to better learning in the long term4. There is a cautionary point here though that is captured by the next pillar: successful.

In order for our ‘activating prior knowledge’ to actually serve its purpose in supporting effective retrieval, we need to set up the conditions for effort for our students, rather than just relaying information at them.

Take these three examples of retrieval for instance:

The third example doesn’t require any cognitive effort from students. It merely states a fact, meaning we have no real evidence that students are even paying attention to it. And even if we can assume that they are, there is no effort involved in the retrieval, meaning it’s missing the elements of ‘effortful reconstruction’ that make retrieval effective.

Of the first two, both could be effortful. So our question moves to, how effortful?

Because if our aim is to keep students in the Goldilocks Zone, we need a measure of that information. Well, that’s where ‘successful’ comes in.

Successful

“A desirably difficult task is one that requires effort but — crucially — results in success.”
- David Didau

The problem with making a task effortful is that effort is only effective if the retrieval is successful. We’re bordering on a dangerous paradox here. This is only made worse by the fact that one core component of gapless instruction is that success must precede struggle. This means we need to ensure success - at the very least, as much as possible.

Luckily, there are a few ways we can set up the conditions for success. The first is rather obvious: teach the content first. Retrieval works by reconstructing things that already exist in one’s memory. This means if we’ve not taught the content before, it’s not a retrieval. It’s a guess.

Of course though, simply having taught something doesn’t guarantee that the students are able to retrieve it. This is where the role of responsive teaching enters the chat.

I hear a lot about teachers using a quick quiz at the start of each lesson as a retrieval activity; however, sending your students off to attempt retrieval questions without monitoring is a risky game. This isn’t to say I wouldn’t use this mechanism, but I would pair it with my formative assessment. That means checking what students know at each point and tempering teaching and prompts to ensure success. Let’s go back to our Poseidon example.

Let’s assume I’ve taught the above to my class in a unit focusing on both Roman and Greek mythology. The classic error here would be to mix up the Roman and Greek gods of the sea - Poseidon and Neptune. In order to ensure success, I would start by asking my students to show me their answers to the above on their mini whiteboards.

If I knew from that the question was too effortful, I might throw out the following prompt:

“Alright, tell me any gods of the sea that you know.”

This questions help me to understand whether students are stuck between Poseidon and Neptune. If that’s the case, I could then temper my teaching further. Rather than telling students the options are Poseidon and Neptune, I would likely throw out a prompt: “Poseidon is a Greek god. How does that change your answers?”

Or maybe the concern is that students couldn’t recall which Greek god was which, in which case I might write up some names and what they’re the god of, and ask students to match the ones they know. My goal is to temper my prompts live based on the information I’m getting from my classes, but ideally still set up the conditions where students need to effortfully reconstruct the relevant information through prompts.

The way I picture this is a bridge between my question and the knowledge sitting somewhere in my students’ long term memory. I think about how I can give them the fewest ‘pieces’ to that bridge so that they can reconstruct it themselves successfully, but where they still need to put in effort. And if they can’t? We reteach it.

Serves Curriculum

One of the core mistakes we make with retrieval practice is that we retrieve a bunch of random disconnected facts at the start of the lesson, but they don’t really feed into what we’re about to teach.

To draw on David Didau again,

“Some of the most useful retrieval prepares students for the lesson they are about to have.”
- David Didau

This means the direct benefit of access questions is that they’re retrieval that is designed to intentionally target exactly what it is you’re about to teach. What I love here is that apart from the very obvious activating prior learning that we’ve discussed, it also builds an additional layer of context to strengthen that knowledge for our students. It means we move them beyond merely knowing facts to knowing where they could be applied.

That actually feeds nicely into my final component around responding to retrieval attempts.

Active Response

Retrieval practice can only ever be as effective as your response to it. There is no benefit in having students attempt retrieval without a response from the teacher, and that same principle applies for access questions.

One core purpose of an access question is about granting access to your new lesson content to students who otherwise may not have been able to access that content; however, if your students attempt the access question and cannot do it, then it’s not really achieving its purpose.

Now there’s two things that could potentially be going on here and to understand these, I’m going to draw on The New Theory of Disuse. This theory challenges the idea that when we learn something and don’t use it for a while, we just forget it. Instead, it distinguishes between two different mechanisms:

1. Storage strength;

2. Retrieval strength.

In short, storage strength is how well learned something is. I like to think about this as being information that is embedded or connected to lots of other pieces of knowledge in our brain, and go back to that image of the cognitive schema of connected it is to other pieces of knowledge in our brain. What’s great is storage strength is generally pretty consistent once we’ve built it up.

Retrieval strength, on the other hand, is not so great. Retrieval strength is how easy that information is to access, but it’s kind of annoying. Think about when you’re at pub trivia and you’re asked a question - perhaps about Rome (the city). You know the answer. But it’s not coming out. It’s on the tip of your tongue. And you know that with just the right prompt you’ll definitely come out with the right answer! You just need that first letter or that hint.

That is an example of a weak retrieval strength. You know the information, but you can’t get to it. I feel the same way about my childhood landline phone number - with a prompt, I can get it, but without that I often struggle.

Why am I telling you this while talking about access questions?

Well, if students don’t know how to do an access question, that could be one of two things: weak storage strength (that is, they haven’t really learnt it) or weak retrieval strength (it’s there, but hard to access). This means part of running our check for understanding is around diagnosing which of these it is. Basically, do they get it with enough prompting? That may suggest they know it. Alternatively, do they not know it at all? Well, that’s probably storage strength.

If it’s a storage strength issue, we likely need to reteach it, and provide opportunities for practice. In the lesson, this looks like lots of similar access questions all right before the new core content. In the example I show below of my slide deck, these access questions are multiple copies of the same question with slightly different numbers (or contexts for my non-maths teachers!) so that I can reteach and continually check for understanding until my students know it.

That means that for the students who were absent or have an instructional gap that is preventing them from accessing the new learning, we are actively filling that gap with our instruction before we move onto the new content. I’ll add in that brief caveat that we’re aiming for 80% or more whole-class success on access questions, noting that sometimes 100% - whilst desirable - just isn’t possible.

Alright, but what if it’s retrieval strength?

If students can’t do something, it might be that it’s just their retrieval strength is weak or tied to different cues. In fact, to really emphasise this, I’m going to refer to the following Tweet by Dylan Wiliam from August this year:

Dylan Wiliam makes the case that if students have really learnt something, they don’t forget that after time. It’s their ability to access it - the retrieval strength - that is impacted. Our way of building that up is by practicing retrieval, hitting on that Goldilocks Zone where that retrieval is both effortful and successful. That’s where our prompting and tempering our instruction comes in.

Over time we’re aiming to reduce our retrieval cues so that the questions we present to our students alone act as retrieval cues.

in summary - what’s an access question??

Please tell me - if even only to humour me, who has spent days over several weeks writing this - that you read everything else to get here.

Here’s a reminder from above:

Before teaching a piece of new content:

• Identify the prerequisite knowledge students need to access the new learning.

• Ask a short question that requires students to retrieve that knowledge immediately before they need it.

• Check the response of the whole class so you know whether students can actually access what comes next.

• Temper your prompts to keep retrieval effortful but successful where the knowledge is there but difficult to retrieve.

• Reteach when necessary & continue to check the responses of the class if the prerequisite knowledge genuinely isn’t there - until they get it!

• Move into the new learning only once you’ve established that students have enough of a foundation to access it.

At the beginning of this post, I said that one of the principles that underpins much of my teaching is pretty simple:

Make success possible.

Access questions are one small, practical way I’ve been trying to actively do that for a while now. It’s nothing new and certainly not something I wasn’t aware of, but by writing my lessons with access questions before any new core content, I’m being a bit more intentional about this in my lesson design. While I can’t control the instructional gaps that have accumulated prior to students joining my class, nor can I control for other factors like genetics or past absences or whatever else, I can control what happens in my classroom. And more to the point, I don’t have to simply accept those knowledge gaps as being evidence of what a student is capable of. Instead, I can believe that my instruction can make a difference and I can teach knowing that.

It means asking what students need to know to access new learning before teaching it. If they do know it, great. That’s awesome retrieval and activating prior knowledge. If they don’t, I have a real opportunity to fill in that gap before it compounds.

I like the phrase ‘access question’, because it reminds me that this isn’t just a bank of random retrieval questions or just another add-on to a crowded lesson. This is a way of making my instruction more responsive to the students sitting in front of me and actually making my lesson land for so many more of them. Because if knowledge builds on knowledge, then one of the most powerful instructional moves we have is making sure our students have that foundational knowledge before we try to build on it.

Share

I’ve said before that I learn so much every time I go to these conferences, and being in rooms filled with people of a similar mindset and dedication to just being the best educators we can be is a pretty great feeling. As my podcast co-host, , said in his keynote:

“If you come from a school where everyone is a bit grumpy, you can come here and meet people who aren’t.”
- Greg Ashman

Or, if you’re like me and work with some incredible educators usually at your school, you can just have a reunion with all your edu-mates from across the country and learn some cool stuff.

In a quick interim post this week (I’ve got another very practical classroom one coming), let’s recap that time I visited the nation’s capital, Canberra, and got to talk education. If you want more, Greg Ashman and myself will be talking about researchED Canberra on the next episode of When Will They Learn? podcast, released this Friday morning at 5:30am AEST.

keynote #1: greg ashman

If you know me, you won’t be surprised to hear that I misjudged my timings a little in the morning while enjoying a hotel buffet breakfast, and basically had to be dragged to get my second coffee for the day, which I insisted on having because The Cupping Room in Canberra is genuinely one of my favourite cafes, and then to the conference. In classic Caiti style, I strolled in barely before the official start time of 9am. And lucky I did, because Greg Ashman decided to do a roll call to me during his keynote, where I was forced to reveal my back corner hiding spot.

Chaotic morning aside, Greg spoke about the ‘what, why and how’ of the science of learning. His keynote began with an explanation of cognitive load theory and gave some background on how we learn. At one point, he said he was going to give us all a simple instruction, before flicking to a slide that said “Don’t think of a white bear”. While dance music and lyrics repeating the phrase blasted across the projector, my mind started to consider whether a panda bear counts as a white bear with black patches or a black bear with white patches. I had creatively failed at the same task as everyone else in the room.

As the session went on, Greg delved into the meaning of explicit teaching:

“When I talk about explicit teaching, it is not the point where you explain something; it is the entire system.”
- Greg Ashman

When people talk about explicit teaching as being a portion of their lesson, I know we’re talking about different things. Explicit instruction includes the process of matching instruction to a learner’s level of expertise, gradually releasing responsibility to learners so that they can complete independent problem solving by themselves. To really emphasise this, Greg broke explicit teaching down into four core components:

• Complex ideas are broken down;

• Each small component is taught and processed before bringing components back together;

• Highly interactive;

• Gradual release from teacher to student.

Those last two points are extremely important. Explicit teaching isn’t reading a beautifully sequenced PowerPoint of slides at students. It’s about opportunities for practice and teacher response. As Greg puts it, “if all you do is display slides regardless of the information you get from children, it’s not very good”. In fact, it’s probably very boring.

That not only builds a case for interactivity, but also responsiveness to inform the gradual release of responsibility from the teacher to the student so that students can do “complicated things” by themselves. This is what can be lost when elements of effective practice are reduced to checklists or just using a certain set of resources. While these individual components matter, they sit within a much larger instructional system and are underpinned by many theories, decisions or professional judgments by the classroom teacher.

Knowing the principles of effective instruction though is only part of the challenge. How do you get this happening consistently across a school?

This was the part of Greg’s session that I took particular interest in. Schoolwide change is a big part of my work with schools and was the theme of the session I presented in the afternoon in Canberra. It’s tricky stuff to roll out on a schoolwide level. As Greg says,

“If you stuff it up, it’ll just be another one of those fads”
- Greg Ashman

Of course the classic tendency is just to launch a bunch of new ideas in January when half of your staff are jetlagged from their holidays, but Greg proposes something different: a Term 4 pilot (halfway through second semester of the school year for my non-Australian readers).

I love love LOVE this suggestion. It overcomes the classic issue where teachers are managing new classes at the start of the year and, as Greg says, often there isn’t the relationship for these classes to be as forgiving if a trying something new doesn’t go to plan. In my opinion, the beginning of the new school year is also the time when teachers are busy trying to plan or prepare for new classes, and giving them another thing to try probably just won’t happen.

Greg also makes the point that this avoids the issue too where we know teachers are tired of new January fads being launched every single new school year. You know exactly what he means. The classic professional learning days at the start of the year where some leaders used their holidays to brainstorm the next flavour of month in instruction, and decide to launch it. It’s usually a significant amount of administrative work that is forgotten by April. The problem is that for many teachers, this has happened so many times that it’s easy to assume that even good ideas will be just another time-wasting administrative task, forgotten about before the next school holidays.

Finally, Greg makes the point that you should pilot with enthusiastic teachers. If you’ve seen me speak about diffusion of innovations, you will know that even the research points to running pilots with your most eager staff. New instructional techniques will never be perfect on the first go and translation into your context is crucial, but might hit a few speed bumps. Rather than piloting a new idea with the staff who want it to fail - because they will make it fail - pick your most eager crew and let the social game do the work for you.

jeanette breen: writing in the curriculum

Jeanette Breen’s session brought together writing and curriculum under a deceptively simple underpinning idea: you can’t write about what you don’t know.

Writing relies heavily on knowledge. It’s not purely about vocabulary and grammar, but about the content of the writing itself. This means that if we want our students to be able to write successfully, we need to deliberately plan and sequence knowledge across the curriculum in such a way that it gives students something to write about.

What I loved in particular about Jeanette’s session was that she unpacked this through discussing the different forms of curriculum that exist. There is of course the recommended curriculum that bodies like ACARA put out and tell us students should learn; however, even translating the recommended curriculum into something both useable and effective is something that introduces workload and variation into our practice. That’s where the next levels of the curriculum come in. There’s the stuff we actually teach, the resources that support it and, ultimately, what the students actually learn and remember.

Drawing on Clare Sealey’s work, Jeanette made the following point:

“The learned curriculum is the only curriculum that actually counts in the end.”
- Clare Sealey

It doesn’t matter how great your resources or teaching are if the knowledge is forgotten. Or, as Clare Sealey puts it, “knowledge cannot empower if it is forgotten”.

This creates an interesting challenge for schools. We have to somehow interpret curriculum guidance and turn that into actual teaching, and then get our students to actually retain that knowledge. That’s a monumental undertaking for schools, and it’s hard to know where to begin, but Jeanette offers a solution to this. She suggests considering what it is students must remember.

This is a good part of the premise behind a knowledge-rich curriculum. It’s about considering what knowledge will inform student writing and interpretations, and considering how to deliberately map that knowledge: both vertically across year levels and horizontally across different subjects in the same year. I often consider this in the sense of learning about historical events in History, before completing a novel study set in that time period and location in English. However, Jeanette doesn’t just suggest doing this. She builds up the case for this work.

Outside of school, we have little control over what is happening for our students. There are differences that exist between students as a result of disadvantage, their upbringing, their exposure to knowledge and their genetics, amongst several other things. Despite this though, we can control what students encounter in the classroom. Jeanette drew on some of my favourite quotes from this, including:

“Ability is the consequence, not the cause, of what children learn.”
- Graham Nuthall

Whilst we can’t control other factors, we can make deliberate choices about how we sequence and structure the curriculum, and the expectations and practice opportunities that support learning within this.

To really drill down on this, Jeanette consistently threaded in examples of the incredible work done at Templestowe Heights Primary School to do exactly this. She showed how the writing journey from Foundation to Year 6 (primary/elementary school) was mapped, with the initial proportion of time being spent largely on spelling, handwriting and grammar/punctuation in Foundation, before moving to a greater proportion of text composition in Year 6. Unsurprisingly, this aligns closely with the work in The Writing Revolution that suggests starting with the most granular details in writing before moving onto text composition.

Importantly, writing instruction doesn’t sit separately from the knowledge accumulating across the curriculum. The two are deliberately woven together. I like this way of thinking where rather than treating each unit as an isolated parcel of content, we instead of intentionally include opportunities to deliberately build on the knowledge they already have and give them knowledge to use later. A knowledge-rich curriculum isn’t about creating enormous lists of disconnected facts for students to memorise, but rather about deliberately planning the knowledge that students will encounter, and how this can support their writing over time.

Because while we can give students all the writing or paragraph structures in the world, the original point still stands: you can’t write about what you don’t know.

brendan lee: different types of learning

This session was NOT about learning styles. I repeat. It was NOT about learning styles. Instead, Brendan spoke about how there are different types of learning required for different purposes. Like many of Brendan’s sessions, he offered a refreshing and novel take on something that is easy to talk past in education.

We know there are different types of learning. For instance, kicking a ball is of course different to memorising one’s multiplication tables. However, we often don’t talk about this, despite this clearly being especially relevant in instructional design.

I’ll offer a brief summary here, but I want to point out that Brendan has got a Substack article where he goes into far more depth than I will.

1. Psychomotor learning

This type of learning is all about a voluntary physical movement. My brain initially when to sports teaching, but Brendan was very quick to point out that this includes other things like learning to use a pencil grip, learning how to pronounce a rolled ‘R’ in languages, or even learning to use a compass to draw a circle in maths.

Even in drama, students needing to imitate the physical cues and posture of an elderly person would be psychomotor learning.

Brendan broke down how to teach psychomotor skills in five steps, but I’ll let you read that directly from his Substack (linked above).

2. Facts

A fact is something a student can recall. This is our classic ‘rote learning’, and includes things like multiplication tables. What Brendan emphasised here though is that the process by which students come to an answer is part of what determines whether it is a ‘fact’ or not.

Take for instance, multiplication tables. A student may just have memorised that 6 multiply 4 is equal to 24, but another student may come to that same answer by drawing arrays. Only the first is fact learning.

This made me reflect a bit on elements of my own teaching from the past week. I’ve needed some of my classes to memorise the fact that p-hat is equal to X divided by n (it’s just a formula for my non-maths readers!). I thought about how I ‘taught’ this by just having my students recite it, quickly and frequently. My focus wasn’t on using the formula so much as it was just being able to recall it.

3. Procedure

Next was teaching a procedure, which is teaching students to follow a fixed sequence. For instance, this could be lighting a Bunsen burner, working with a known algorithm to answer or problem, or even follow the steps of structures like ‘PEEL’ for writing paragraphs.

Again, Brendan’s Substack sets out how to teach these, but what I loved here was he referenced identifying prerequisite facts and skills as being part of the teaching process. This is something that I’ll talk to more in my next post for this week.

4. Concept

A concept is something that can be classified based on defining features. Think about the classic example of identifying what shapes are or aren’t triangles. Teaching a concept involves considering what features of the concepts are always the same, what are allowed to change (for example, the colour or orientation of the triangle), and then juxtaposing both ‘close-in’ (similar) non-examples, and maximally different (as different as possible) examples.

This is the sort of learning where you need a student to classify an unseen example or stimulus using its features. While my brain thought about shapes mostly, Brendan showed examples of artwork, and how knowing about colours, subjects or brush strokes can reveal the ‘type’ of art - impressionist, pop art, etc. I’m not an art teacher, nor expert, so please excuse any language misuse there.

5. Principle

Put simply, a principle is something that applies an “if-then” relationship, and is focused on applying a rule to a different situation. This could be applying existing rules to new circumstances, such as applying rules of grammar (e.g. subject-verb agreement) to unfamiliar sentences, or even using mathematical properties in different scenarios.

Again, my brain went to the distributive property in mathematics. Funnily, Brendan references this one in his Substack. For those of you who kinda remember maths, but not totally, this is colloquially just “expanding brackets”. I recall teaching this essentially using these ideas, without being able to name them. Often students conflate 3(2 + x) and 3(2)(x), and try to distribute the 3 in both scenarios. This would be an example of a principle, in that the distributive property is only applied in the first example where there are multiple terms inside the brackets, but not in the second example.

6. Strategy

Strategies are how we begin to move our teaching into the very unfamiliar problem territory. This is about problem-solving - something that does sit within the explicit instruction framework. What I love is that Brendan draws on the instructional hierarchy here to get into how open-ended problems can still be taught and students can be provided with the skills and ‘strategies’ needed to give them success on these problems. The emphasis here is very much on making sure they having the prerequisite knowledge first, of course.

This is all about how to teach for the generalisation and adaptation stages within the instructional hierarchy.

Summary

If any of that interested you, which I would imagine it would, just a reminder that Brendan’s detailed post on that is available here, where he literally breaks down each of the above into 5 steps for teaching them.

sarah siakew: the implementation gap

Next up was Sarah Siakew, who I’ve been fortunate enough to chat all things research with, and was lucky to see in person. Sarah spoke about the implementation gap: we invest heavily in education, we have a solid evidence base, so why does this not translate into practice?

Regardless of what you consider to be the ‘correct’ evidence base, this session spoke to a deeper question: what stops change from being implemented in education?

An approach can be well-supported, a school can commit to implementing it and it can still somehow fail to be implemented in the classroom. Sarah speaks about how as the evidence around practice travels through a system, it encounters friction at each different transition point. Even the movement from school leaders to heads of faculty to classroom teachers introduces at least two transitions where that information can be diluted or changed.

Even still, the practice we see in the classroom is also only a very minor snapshot of exactly what has been implemented. Sarah used the analogy of a swan on water, emphasising that focusing only on what we can observe in a lesson misses everything beneath it: the teacher knowledge, professional learning, routines and the several hundred decisions that teacher is making every moment. This is why purely copying an instructional technique without an understanding of the underlying factors can lead to lethal mutations. Its inclusion becomes performative - basically a checklist of things to do to get a ‘good score’ on a lesson observation.

As Sarah so eloquently put it, “any instructional technique is only as good as the knowledge that sits behind it”. This leads to a direct action in shifting practice: professional learning needs to therefore do more than simply introduce teachers to strategies. It needs to deepen the knowledge underpinning why they work.

Of course though, Sarah points out that professional learning alone isn’t sufficient. That is something I wish I’d been told a couple of years ago. Even when staff believe something is worthwhile, it doesn’t mean it will happen. Sarah compares this to knowing that eating healthy and exercising is ‘good for us’ - knowing that and putting in the effort to do it are two totally different things.

So what advice does Sarah have for us? Well, she posed a question that I think is terrific way of framing a lot of the work in implementation:

“What would have to be true in this school for this practice to be fully embedded?”
- Sarah Siakew

This is really about identifying barriers to implementation, and then determining enablers. These are unique to each school, but is a useful framework for considering schoolwide change. It - rightfully - shifts the conversation away from simply examining classroom instruction, and instead towards examining the conditions that allow them to actually implement that instruction well.

me: beginning schoolwide change

At researchED, there are multiple sessions running simultaneously, and you have to pick just one to attend. This is already challenging, and I 100% have friends who split ourselves between sessions so that we can cram in as much simultaneous ‘learning’ as possible. I crack the joke each time that I think I’ll skip going to my own session and instead see someone else scheduled on at the same time, because I already know what I’m going to say. I can’t tell you how true that was this time round, because the group blocked on at the same time as me were all incredible.

I’ve run some version of my ‘The First Five Steps: Beginning a Schoolwide Pedagogical Change’ a few times across the country now (3 states), and it’s officially retired now!

As someone who began as a maths instruction presenter, then moved into general science of learning, I’ve enjoyed presenting on implementation science. I work heavily in this space with schools aiming to shift their entire pedagogical framework to an evidence-informed model, and in this work, I found that the hardest part isn’t telling schools how to change their pedagogy. Instead, the tricky part is change management. This session came from me attending every single school change presentation possible when I first got into this space, only to find that they were far too high level for me to walk away with a set of actions that I could apply tomorrow. Hence, my ‘First Five Steps’ was born: a model that takes the research, breaks it into steps, gives specific leadership actions within each, and aims to be both general enough that everyone can apply it, but specific enough to be useful. Difficult aim, by the way. Don’t recommend that.

Anyway, the overarching summary slide is below - complete with my signature absence of capital letters, despite running this presentation in the nation’s capital city.

Of course, my entire presentation was framed around the following:

“Expertise does not exempt us from learning. If anything, it increases our responsibility to keep refinining.”

Doing something well doesn’t mean there isn’t more to learn. I know speaking about ‘fidelity’ isn’t necessarily popular, but I like it because I feel it emphasises that you can always go deeper into something. You can always learn more to make an existing strategy more effective, or get better at something you already know.

That’s why I love working with schools around Australia on professional learning and schoolwide change. There’s something very empowering about getting to work with the mindset that improvement isn’t about throwing out everything you’ve been doing. It’s about just wanting to learn more about it.

Of course, if you’ve missed this session, I will be running a different take on practical implementation science as part of my 2027 conference run, so stay tuned for my ‘silver bullet’ presentation (full disclaimer: there isn’t a quick easy ‘silver bullet’, but there are a lot of strategies that can work together to practically shift practice).

keynote #2: jenny donovan

Now we know I am a HUGE AERO fan, and seeing Jenny Donovan present was something I’d excitedly waited for all day. Unsurprisingly, she did not disappoint.

Unsurprisingly, her presentation was ‘clever’ in every sense: beautifully structured, with a carefully sequenced thread running throughout, and equally sharp in its content. It was the exact sort of presentation you would expect from someone who has contributed so much to Australian education. As Tweeted, “Let the record show that Jenny Donovan is one of the best things to happen in Australian education for decades”.

The session focused on writing. One of the ideas that stuck with me was the complexity of writing itself. It’s far more than purely putting thoughts on a page or knowing that a sentence ends in a full-stop/period. It requires students to coordinate knowledge, grammar, motor skills and the demands of genre, all whilst managing the known limitations of working memory. And unlike spoken language, many of these things cannot be innately acquired by students.

If there is one core takeaway from this session, it would be this: We need to teach students to write - explicitly and systematically.

Why? This is at least partially because writing is how students commonly express what it is they know in different subjects. But also being there is something inherently valuable - and genuinely beautiful - about being able to write. I think the latter should really be emphasised.

Recently I was chatting to a teacher about why we teach certain things. We agreed that there are probably significant amounts of the curriculum that have very little bearing on students’ future lives and many things they will never ‘use’ again. But sometimes, it is simply nice to know things.

I feel the same way about writing. Perhaps it’s simply nice to just be able to write. Jenny opened her presentation with a story about love letters that she and her partner had written to one another many years ago. It was a lovely way into a presentation about writing instruction (especially for my hopelessly romantic self who was intently listening), but it also drew my mind to how special the simple privilege of being able to write is. I feel lucky to be able to write to communicate. That’s about finding the right words to say something or knowing the grammatical rules enough to choose to bend them deliberately. It’s taking ideas that exist somewhere in my head and turning that into something that another person can read and, somehow, understand. There is something quite extraordinary about that.

Jenny tied this idea into the intentional use of words - yet another witty way to weave this point into an extraordinarily well-worded presentation. For instance, she spoke about the problematic use of the word ‘innovation’ (something I’m sure I’ve done):

“An innovation has to be good to be considered an innovation, but you can’t prove an innovation is good until you do it”
- Jenny Donovan

Of course, in schools, writing also has a much more pragmatic function. It is one of the principle ways that students demonstrate what they know. Writing is more than just having something worth saying - it’s also about knowing how to say it appropriately. And the challenge with that? The rules change based on the subject or format.

The way that a student writes an evaluative response to a piece of literature is different to how they would evaluate a scientific experiment or a source in history. That’s not even to mention the differences in vocabulary that exists across different subjects. This means that writing instruction cannot sit solely with English teachers. If we want to support our students in writing, we need to weave this throughout the curriculum - just as Jeanette Breen emphasised in her session.

If you’re here wondering how you would even begin to do this - particularly if, like me, you don’t come from a writing instruction background - you don’t need to worry. After all, this is Jenny Donovan we’re talking about. In keeping with what AERO has continually sought to do for Australian teachers, Jenny didn’t simply make the case that writing needs to be taught explicitly and systematically and leave us to figure out the rest. She pointed to a number of practical resources that schools and teachers can draw upon to actually do it:

• School writing instruction framework | Australian Education Research Organisation ← this one contains three professional learning courses that I will be jumping to complete as quickly as possible!

• The Writing Revolution - Empower teachers to help students

• Couch to 5k Writing - David Didau: The Learning Spy

And those aren’t tools just limited to English teachers. Again, I own both of the two latter books, and am about to complete AERO’s courses. I’m also a maths teacher.

To quote Jenny herself in what has to be my favourite quote from the day:

“As a research organisation, I don’t just want to be throwing dead cats on the table.
Don’t get me wrong; it’s fun. But I want to offer solutions”
- Jenny Donovan

Because that is exactly what AERO, under Jenny’s leadership, does. They provide suite after suite of practical resources that turn evidence into practice. For instance, AERO’s work uncovered three central problems in writing instruction: insufficient time dedicated to writing instruction, insufficient frequency of practice and insufficient pedagogical knowledge. The nice part is that this isn’t a problem without an answer.

Whilst there is no ‘science of writing’ in the same way there is reading, there is a solid evidence base around what effective writing instruction looks like. This is about breaking complex writing down into smaller components. It’s about practicing at the sentence level over and over, and ensuring students secure one component before they move onto the next. Or, as Jenny emphasised, go slow early to go fast later.

Perhaps my favourite part about this session was exactly this. Writing is extraordinarily complex, but that doesn’t mean effective writing instruction needs to be mysterious. We have a solid evidence base for how to teach students to write, and we can use that to do just that.

And given just how extraordinary it is to be able to turn something in your head into prose that another person can understand and grasp meaning from, I think that’s something worth teaching well.

a summary

As always, this conference was one of my favourites. It never ceases to amaze me that a grassroots movement built around getting research into classrooms in a way that is genuinely accessible to teachers has made its way all the way here to Australia.

Perhaps what’s even more remarkable is the number of teachers who willingly give up their Saturday to attend - or those who travel from across the country just to do so. This isn’t simply a day to meet the required number of annual professional learning hours. This is about educators who just want to learn more about how they can be the best for their students.

There’s something pretty special about being in a room full of people who care that much about education.

Share

Anyone who has been stuck spending time with me in the past week has also quite likely also been stuck hearing all about the Australian Centre for the Public Understanding of Education (ACPUE). Quite possibly at length.

As it would seem, I have become a little too invested in this whole thing. It’s like a real-life Agatha Cristie style mystery, where at the end they’ll reveal that the people behind ACPUE is everyone, or something like that1. It’s featured in my education group chats2, and even those in my personal life who have absolutely zero reason to care about the internecine disputes of Australian education have found themselves receiving detailed briefings on the ACPUE saga.

At one point this week, I even found myself explaining the whole thing to a doctor3, who had absolutely no reason to need to know about whatever factions or mysteries are forming in the edu-world. To his credit, he did a remarkable job of following along with something that is unlikely to impact his life outside of hearing me mention it.

Eventually, he asked why some people are so resistant to evidence-informed practice in education - the exact sort of question I suppose you could expect from someone working in a profession that does use evidence-informed practice. This was both an entirely reasonable question and proof that my explanation of this story had started several chapters too late.

From the outside, this whole thing is a bit strange. There’s a mystery organisation that has popped up, analysing education research, called ACPUE (“an unfortunate acronym” to quote Greg Ashman). But we don’t know who they are. And they’ve released a report/analysis on an AERO (I’m breaking academic writing rules by choosing to introduce AERO later) report. And that analysis claims that, amongst other things, “42% of citation checks” in AERO’s report don’t match. And now people are discussing it on Twitter/X, LinkedIn, Facebook and here on Substack, and I’m turning it into a real-life murder mystery dinner party.

So, let’s start at the beginning.

What’s the deal with ACPUE?

chapter 1: evidence-informed practice, the short story

I am going to keep this chapter as brief as I can, while still giving you enough context to understand what’s going on. I will start though - as anyone aiming to report on something transparently should4 - by declaring that I do have a bit of a personal bias here. Given you’ve made it to my Substack, you likely know I am a fan of evidence-informed practice, specifically science of learning and cognitive science. I work heavily in this space, present on evidence-informed practice, conduct research into cognitive science, and consult with schools in this space. I think that’s an important framing, as it will impact my interpretation of these events and I’d like readers to be able to make up your own mind on how I’m reporting about this.

In short, there have been some reforms to Australian education that have brought the idea of evidence-informed practice (in this case, I’m defining this as science of learning evidence) into common practice. Without going into too much detail, a few examples include:

• The states of Victoria and New South Wales both advocating for evidence-informed practice through both the VTLM 2.0 and NSW Explicit Teaching focus;

• Australian universities must embed “mandatory” core content in order to be accredited to provide initial teacher education (teaching degrees), which comprises research around how our brains learn and how this informs instructional techniques, such as explicit instruction;

• The proposal to establish the Teaching and Learning Commission in Australia, which would combine the works and responsibilities of overarching education bodies in Australia, including those in charge of the curriculum and AERO (who I will talk about later, but in the meantime, all you need to know is they also draw on that same body of evidence I’ve been referencing so far in this post).

Whether you like evidence-informed practice or not, it’s really at the forefront of a lot of educational movements within Australia5. And for anyone who doesn’t like evidence-informed practice, this is likely a pretty unsettling time.

But to go back to the question I was asked this week: why would people be so resistant to evidence-informed practice?

Well, there are a lot of reasons, and I’m very conscious not to dismiss or minimise the very real reasons people may have for that resistance. To briefly list a couple - noting very clearly that a brief list cannot possibly do the complexity of these reasons justice - these reasons could include6:

• Teaching is tied very closely to identity, and a change in practice can be, for many reasons, a huge shift in this identity;

• The fear of losing autonomy - in a variety of forms - in the classroom;

• Different beliefs about teaching, such as beliefs in learning styles, inquiry learning, differentiation or hands-on learning;

• Poor implementations of evidence-informed practice or low support enacting new practice, making it hard to contextualise in different contexts;

• Some things that look and feel like learning aren’t particularly effective, and the things that are effective don’t always feel that way;

• There is a lack of trust in education movements, stemming from years of useless fads that were nothing more than a time wasting exercise;

• Some people just sit on a different side of the research, where they follow different lines of evidence;

• And I’m sure there’s some sunk-cost fallacy here where educators have dedicated hours and hours into resources that they might not be able to use anymore.

None of that is intended to dismiss or oversimplify why people may be reluctant to enact evidence-informed practice. The list is neither exhaustive nor detailed enough to properly acknowledge the very real factors that can make teachers reluctant to move towards it. The important takeaway is there’s a lot of reasons that this nationwide movement towards evidence-informed practice might be a bit unsettling for them.

chapter 2: aero and the report

Now let’s introduce AERO, the Australian Education Research Organisation. Who are AERO? I’ll let them introduce themselves:

“AERO is a ministerial-owned company governed by a Board. We are jointly funded by the Commonwealth, state and territory governments. We conduct research and share knowledge, positioning Australia's educators at the forefront of education research and promoting better educational outcomes for Australian children and young people.”
- AERO

In short, AERO takes education research and distills this into practical reports, guides, videos, resources and basically everything else helpful, so that Australian teachers can actually use this evidence in the classroom. At this point, in the spirit of being transparent about any personal biases that may be present in my reporting7, I will mention that I am a strong supporter of the work of AERO.

Why am I Team AERO all the way?

It’s actually pretty straightforward - AERO fills an important gap between education research and classroom practice. Teachers are busy, the research literature is enormous and often inaccessible, and most of us don’t have the time or ability to independently interrogate a heap of studies relevant to our practice. AERO synthesises that evidence, identifies useful ‘best bets’ from this, and translates them into practical resources that teachers and school leaders can actually use (please see some linked here if that sounds great to you). That doesn’t necessarily mean teachers should accept anything AERO publishes uncritically, but it’s an accessible, useful starting point from which teachers can exercise their own professional judgment.

Yeah but was does AERO have to do with ACPUE?

We’ll get to that. First, what you need to know is that not everyone is quite as much of an AERO fan as I am. And, given everything we’ve just covered about the tensions surrounding evidence-informed practice in education, it probably shouldn’t be surprising that AERO has attracted its fair share of critics.

Those criticisms aren’t particularly new. But in 2025, they suddenly had a focal point.

Australia’s Federal Education Minister Jason Clare commissioned the consulting firm, KPMG, to undertake an independent performance evaluation of AERO. Basically, it was meant to consider its effectiveness, operations and, basically, whether it was doing the job it had been created to do. As part of this review, people were invited to contribute.

As this review was underway, a series of increasingly critical pieces about AERO began appearing online (such as this one), often questioning the evidence base that AERO draws upon. And of course, this meant those of us who like AERO also joined in. What began as a government-commissioned review of a research organisation had become something of a proxy war over evidence-informed practice itself.

What was the outcome of the report? And what about ACPUE?

Well, here’s the big sticking point: the report hasn’t been released publicly.

KPMG completed its review and provided its findings to government, but despite requests for its release, the actual report remains out of public view. So we don’t know exactly what KPMG concluded about AERO, what evidence it relied upon, how it weighed the various stakeholder contributions, or what recommendations it ultimately made.

Which is a little inconvenient when everyone spent so much time fighting about it.

If you’re following my story so far, you also won’t be surprised to know that people aren’t happy about the fact this report hasn’t been released to the general public. I won’t speculate here about what is or isn’t in the report, but if it were released, I would love to get my hands on it. Not because I’m predicting what this apparently top-secret report says, but because I suspect I’d have quite a good time doing a deep dive into it: interrogating the methodology, following the evidence, and seeing whether its conclusions actually hold up8.

It’s worth remembering here that KPMG is a consulting firm, not an independent research body. They were paid by the government to answer a particular brief, using an agreed scope and methodology, for a particular client - a client that was paying them to do this report. That doesn’t automatically make their conclusions good, bad or otherwise; however, those incentives and constraints do matter greatly when evaluating any piece of research. Which, as you’ll soon discover, becomes rather relevant to this story.

chapter 3: and then there was acpue

This is where our mystery starts to get very interesting. Because whilst the KPMG report itself has remained behind closed doors, another report evaluating the work of AERO has appeared publicly.

And the organisation behind that report?

ACPUE: The Australian Centre for the Public Understanding of Education.

Haven’t heard of them before?

Don’t worry. Neither had basically anyone else until sometime in the last week or so.

Rather suddenly, this new organisation appears on the scene with a new website and an entire report scrutinising some of AERO’s work.

And yet, for an organisation calling itself the Australian Centre for the Public Understanding of Education, there was one rather curious omission.

Who, exactly, is behind the Australian Centre for the Public Understanding of Education?

chapter 4: acpue & aero, the acronym crossover

Before we get into the ‘who’ is behind ACPUE, it’s probably worth talking about the AERO piece here.

At the time of writing, ACPUE lists several projects on their website, but its first completed piece of work is an analysis of AERO’s 2023 How students learn best paper.

What might be worth noting, however, is what the other projects are that are listed on ACPUE’s website:

If you are at all involved in evidence-informed practice, you might look at the above and notice that despite ACPUE’s claims that they “are not against evidence” (ACPUE Website), their choice in upcoming projects certainly seems to have a target on evidence-informed practice. Recall me mentioning VTLM 2.0 earlier in this post? Or the core content in initial teacher education degrees I mentioned earlier, which is covered in the TEEP report? Cognitive load theory? Or my favourite, randomised controlled trials???

No really. Are we actually going to review RCTs?? Seriously??

So, without knowing who is behind ACPUE, I feel we’re getting a sense of their agenda perhaps.

Okay, but what does their AERO project say??

Essentially, ACPUE went through the claims made by AERO in their 2023 How Students Learn Best paper (really great read, by the way), classified each citation as either ‘matches’, ‘plausible’ or ‘does not match’. From this, it arrived at some pretty damning headline figures: 47% of claims did not match, 42% of citation checks did not match, and 67% of references were misused at least once.

Which, if accurate, would be a pretty significant problem for an organisation whose entire purpose is evidence.

There is, however, a problem.

I would love to go into this and pull apart their methodology; however, I don’t have to. My podcast co-host, Greg Ashman, already did this in a terrific post here. Whilst Greg acknowledges some citation issues are genuine, there are some other issues with these headline figures:

“I cannot reverse-engineer these figures. The table lists all the citation checks and does not give results by claim. The citation checks show that 40.7% supposedly ‘do not match’ rather than 42%. And like the claims, there is no data on references being ‘misused’.”
- Greg Ashman

Likewise, he found that quite a few of what ACPUE have suggested are citation mismatches seem to be differences in interpretation, or opinion, and not evidence that AERO’s claims are unsupported.

There’s clearly a few issues here. Any good analysis should be sufficiently transparent in its methodology that someone else could follow it and produce the same results - particularly in this style of analysis. At present, at least from the information publicly provided, Greg couldn’t. But I digress. A methodological rabbit hole isn’t the mystery I’m interested in here.

I want to know who did this analysis.

chapter 5: we’re going to need a bigger corkboard and some more red string

Especially if we want to find out who is ACPUE.

Don’t even worry, because Greg Ashman is on the case, in what is shaping up to be one of my favourite education detective stories of all time9.

After finding no names of a CEO, board, researchers or really any identifiable humans at all on ACPUE’s website, Greg did the obvious thing: he emailed them and asked who they were.

Four days later: no response. So, naturally, he started digging himself.

And this is where we’re going to need a bigger corkboard and significantly more red string10.

Again, Greg has written this terrific post outlining his searches and deep dives to find out who is behind ACPUE.

But you might be thinking, Caiti, why do you even care about ACPUE??

Here’s why. For an organisation that claims they “are committed to rigorous, open inquiry” (ACPUE website), it’s curious that they don’t apply those same standards to themselves. As I’ve stated numerous times in this post, declaring one’s personal interests is an important part of maintaining integrity in research. It means that anyone reading claims that identify any potential conflicts of interest, sources of bias11, and other incentives, when they’re evaluating those claims. That’s something that cannot be ascertained when no authors behind the report are listed. And, honestly, I’d say it’s a little hypocritical to critique standards of rigour and transparency in the work of organisations like AERO that ACPUE is unwilling to apply to itself.

Alright, who is ACPUE?

Well, Greg’s digging turned up a few names, so I’ll let you read his post to find those, because I think reading the full story is important. Greg doesn’t state that anyone is behind ACPUE, but presents any relevant findings, and the full context is important there.

But I do have an update.

chapter 6: the time caiti was up writing a post and taking screenshots and accidentally got at least a partial answer to this mystery, as if the universe knew this story needed an ending

In my casual trawling of the ACPUE site for screenshots tonight, I came across an addition that was not there a few hours ago:

You can bet I immediately contacted my group chats, Greg and the doctor from the story above when I saw this one.

David Clements and Nataša Ciabatti now have names attached to the ACPUE website.

Now, I want to be careful here. Their names appearing on the site does not necessarily tell us that they founded ACPUE, authored its AERO analysis, run the organisation, or are responsible for any of the things I’ve discussed above. They might not even know they’ve been named. I’m certainly not going to attribute work to either of them that ACPUE itself has not attributed to them.

But it does mean that, after an entire evening spent asking who the actual humans associated with this organisation are, we suddenly have two actual humans who are quite possibly associated with this organisation.

At approximately the same time I happened to be writing a murder-mystery-themed blog post about finding them.

And now you’re up to date on ACPUE and AERO and whatever other acronyms I’ve used.

is there a sequel??

Probably. I’ve only done a brief skim of the research outputs of both people named above, because - honestly - I’m tired, and I wasn’t expecting the mystery to start solving itself while I was still writing the post.

There are certainly some interesting threads there that I’d like to follow, particularly given everything we’ve just covered about AERO, evidence-informed practice and the importance of being transparent about the perspectives and interests we bring to our analysis.

But that is a future me job - or more than likely, a future Greg job, since he is a far better detective than myself (I really just jumped in to update everyone on the story since I’m finding myself explaining it multiple times a day every day). I have a sneaking suspicion that my corkboard and red string won’t be going anywhere just yet on this case.

Share

• NAPLAN = National Assessment Program - Literacy and Numeracy

• It’s an Australian standardised test that students sit in Years 3, 5, 7 and 9 (so around 8-9 years of age to 14-15 years of age).

• Students are tested, across 4 separate tests, on Writing, Reading, Conventions of Language (grammar and punctuation), and Numeracy.

• It’s big news because national results have been released just now!

Unsurprisingly, the conversations around NAPLAN aren’t always favourable. There’s the classic dislike towards standardised testing in general, fears over results, and a newer train of hate features some form of “explicit instruction doesn’t work because the results haven’t magically improved after one or two years of some schools implementing it, some schools sort-of implementing it, and some schools saying they’re implementing it while doing something else entirely”. Did you not realise that system-wide change, at a national level, intended to compensate for decades of instructional casualties, was meant to work that quickly? I actually see how that’s on me. *shrugs in sarcasm*.

Before we come at me and say that it’s been more than 1-2 years, let’s have a look at the state of Victoria, whose ‘VTLM 2.0’ embeds evidence-informed practice, including explicit instruction:

“The department has developed guidance materials and resources to support schools to implement the VTLM 2.0 from 2025.” - Victorian Teaching and Learning Model 2.0: Policy | education.vic.gov.au

Likewise, New South Wales’ explicit teaching mandate for state/public schools only took place part-way through 2024. As a Queenslander myself, I don’t want to talk about us1.

To take a few steps back, we should ask why there’s even a need to pass blame when it comes to NAPLAN. To provide some context, the incredible Pamela Snow summarises our national results as follows:

“The short story on NAPLAN is that it shows a persisting pattern of under-achievement in literacy and numeracy by about a third of Australian students.”
- Pamela Snow

And just to really highlight how concerning this situation is, Snow goes on to say:

“To put this into real terms 132,857 of the 1.3 million children in Years 3, 5, 7 and 9 not keeping up with basic classroom learning.”
- Pamela Snow

I completely agree that this is cause for concern and highlights a need for more conversation about the Australian educational system. I’m glad this headline is being plastered across media outlets and that a quick search of ‘NAPLAN’ on social media is drawing attention to figures that are frankly indefensible. They point to a far larger issue around education, particularly for our least advantaged students.

Yet, rather than focusing on that issue, some social media critics of explicit instruction have instead chosen to weaponise persistent student underachievement in an attempt to take down explicit instruction. Like usual, I’m back here staging a takedown of this takedown, and like I usually do, I won’t link to any specific posts, because my focus is on taking down an argument, not an individual. Although if you are interested, a quick Google search should help.

no, naplan isn’t evidence that explicit instruction is “working”

You read that right. It does sound like an interesting take from someone who spends a good portion of my life working with schools on implementing effective explicit instruction, so let’s circle back to this claim, especially since it feels neither side of the explicit instruction fence is arguing it.

On August 1st, the Minister for Education, the Hon. Jason Clare, spoke at The Garma Festival. In his keynote address, he stated the following with regards to educational funding systems to support students of families all across Australia from a variety of backgrounds, including First Nations peoples:

“It’s also got to be built on evidence.

Evidence of what works.

We know what works.

You will see that next week when the NAPLAN results come out.”
- Jason Clare

Clare clarifies that he is referring to numeracy results specifically here. Results that are, for reference, the highest they’ve been for Years 5, 7 and 9 since 2023 when the new NAPLAN system2 was introduced. Critics have essentially suggested that this is an overreach on Clare’s part, attributing improvements in numeracy to ‘evidence’ - which is likely to mean ‘explicit instruction’, in this case. I don’t disagree with them on this front. Unfortunately, I don’t think we can attribute our nationwide numeracy results to any potential push for explicit instruction, such as those in Victoria and New South Wales that I mentioned above. I’ll get into my rationale for that later.

Although, where I seem to differ from critics is on the next part. Clare goes on to say:

“In schools across the country that have used explicit teaching methods for a while, what’s sometimes called phonics, and small group tutoring, where you give kids who are struggling extra intensive help, we are seeing big leaps.”
- Jason Clare

Taking this out of context, critics have come down on Clare, suggesting that perhaps he’s taking a very narrow look at the data, being that the percentage of students meeting literacy benchmarks has largely decreased since 2023. They are claiming that Clare is ignoring the fact the data suggests the exact opposite of this. In fact, some online critics have gone so far as to say that the above data suggests that explicit instruction is causing the decrease in literacy results.

Ah yes, causality for me, but not for thee.

What I mean by that is it’s an interesting approach to state that Clare is overreaching by claiming that the improved numeracy results have been caused by explicit instruction, but that the decreasing literacy results have been caused by explicit instruction. Which one is it?

However, I think there’s far more to it than this. Clare’s comment above doesn’t, in my opinion at least, suggest that nationwide literacy results are the direct result of explicit instruction. In fact, I’d say he’s very clear about this being limited to “schools across the country that have used explicit teaching methods for a while” (being that he says exactly that in the quote above). Actually, if I were to take it further, I’d point out that in the next line, Clare literally names Highvale Primary School as being an example of a school that has improved literacy results due to explicit phonics instruction.

So while I’m in agreement that the suggestion that the improvements in nationwide numeracy scores are the result of explicit instruction is misplaced (adding in the caveat that I’m not entirely convinced that this is what Clare meant with his comments), I definitely don’t think Clare conveniently chose to ignore literacy results. That being mostly for the fact that in barely the next second, he literally speaks to literacy results at specific schools.

yeah but why are the results not getting better if we’re implementing explicit instruction?

The answer to this is two-fold. Firstly, there’s the matter of time that I mentioned above and will briefly delve into again, but before I do, there’s also the second issue - our implementation.

As mentioned above, Victoria’s VTLM 2.0 only moved into implementation from 2025, which gives it all of around 12 months to supposedly have produced significant change at a nationwide level. Likewise, New South Wales’ explicit teaching push only came into effect midway through 2024, giving it less than 2 years to have done the same. For those of you not from all Australia, or the Australians who forget that there are states outside of the south-eastern portion of the mainland3, these make up only two of thes six states and three territories (I am counting Jervis Bay here too).

Whilst they might be the most populous, it’s worth acknowledging the fact that the other states do not have the same level of push for explicit instruction; however, they do participate in NAPLAN. Just to be very clear around the figures here, that means around 40% of the Australian population4 lives in states or territories without a push for explicit instruction.

Now let’s assume that magically, the 60% of the nation that live in states that are directly pushing for explicit instruction should be able to significantly shift results on their own, and that some of the other schools or teachers in other regions in Australia might also be implementing this practice. If that’s the case, there has still been under two years since the New South Wales’ explicit instruction ‘mandate’ or the Victorian VTLM 2.0 were put into effect. What that means is a Year 9 student in New South Wales has had 6-7 years of schooling potentially without explicit instruction. That could mean 6-7 years of an education that has left significant gaps in their knowledge, and we’re expecting that we can totally turn that around in under two years for all students. Really?? Even our Year 3 students started Prep/Foundation before any of the statewide pushes for explicit instruction. We’ve not even had one cohort make it to Year 3 (the first NAPLAN test) completely under this model, let alone Years 5, 7 and 9.

While we might not like it, the date a reform appears in a policy document isn’t the date we can start a stopwatch on national student outcomes. Large-scale change takes time to work, as inconvenient - and outright devastating - as that may be. In the words of the Shadow Minister for Education at researchED Melbourne5 this weekend:

“My issue with explicit teaching is that we didn’t implement it earlier.”
- Brad Rowswell, MP

He’s right. That would have made it far easier to judge the impact of this instructional technique at a system-wide level.

But let’s move now into my second issue: implementation.

Above, I suggested that all students in Victoria and New South Wales - at least - were receiving access to explicit instruction. It’s a shame that’s not the case.

Brad Rowswell MP noted this as well at researchED Melbourne, when he made the following comment about explicit instruction, or specifically, Victoria’s VTLM 2.0:

“The model is only as good as the implementation.”
- Brad Rowswell, MP

If that isn’t the truth. Yeah, I know. Fans of explicit instruction always say that it’s less effective - or not effective - if it’s not done well. Funny that. Would you believe that poor implementation is a remarkably reliable way to make literally anything fail?

If you feel like ruining your day/night, take a few moments to trawl online teaching groups in Australia for either Victoria or New South Wales, and you’ll see that there are a lot of teachers vocally refusing to implement explicit instruction or evidence-informed practice. Add to this the fact that some people are also implementing it unintentionally poorly, and that’s not really a recipe for success.

To be completely clear, this is not saying that teachers trying to implement explicit instruction are doing a bad job. Frankly, effective implementation is difficult, and the elements that make it effective can often be accidentally undermined in an attempt to execute too many new processes without understanding or having the appropriate supports to experience success. Incorrect definitions of explicit instruction and lethal mutations just add to this problem, meaning that even if everyone in only two states were genuinely trying to implement it, it’s going to take time for the teachers to learn this practice.

Are we catching what I’m throwing?

In plain terms, the 2026 NAPLAN results cannot possibly tell us anything about the impact of statewide pushes for explicit instruction.

Takes a little longer than that to turn a ship one would think.

the curious case of leaders making evidence-informed decisions

Some commentary around NAPLAN results has challenged school leaders who have spent the past couple of years actively restructuring their literacy - and presumably, numeracy - programs to consider the use of evidence-informed practice. For whatever reason, those criticising explicit instruction are suggesting that school leaders need to take a good look at what their data is actually saying, questioning whether they’re really looking at their NAPLAN results or merely looking for evidence to confirm that they were right all along.

Ignoring the fact that it is extremely dismissive - and frankly, insulting - to suggest that school leaders, certainly the good ones, don’t examine their NAPLAN results in detail, there is another possibility being overlooked: perhaps this data is precisely why schools adopted explicit literacy and numeracy programs in the first place. For many leaders, looking at the evidence was the catalyst6 for the move to programs that have been shown to work.

For example, let’s look at Djidi Djidi Aboriginal School in Western Australia, who were recently reported on by ABC News for their progress in early reading. The article states:

“Four years ago, testing at Djidi Djidi Aboriginal School revealed 16 of the 18 students in years 5 and 6 were at or below year 3 level in literacy.”
- Conor Duffy & Cath McAloon, ABC News

And what did they do? Well, the principal, Karen Augustson, attribute the success to the following:

“The principal attributes the success to a combination of the latest evidence-based teaching called explicit instruction, a trauma-informed practice known as the Berry Street Education Model and an embrace of Indigenous culture.”
- Conor Duffy & Cath McAloon, ABC News

If you’re looking for the evidence though, please see below (taken from the same article):

The above image shows the improvement of Djidi Djidi in reading from 2024 to 2025 under their new model, as compared to ‘similar schools’ (defined above).

Now before we tell me that I’ve cheated and drawn on outdated 2025 data, I know. It’s because 2026 results have just been released to schools. Schools have likely only just started the deep dive sifting through this data (a thing we know good leaders take time to do), and it’s certainly not publicly available for me to draw upon here. Conveniently, that doesn’t really matter for this argument anyway.

Similarly, I’d like to point out that Djidi Djidi isn’t the only school whose results have shifted when they transitioned to explicit instruction models. Let’s talk about Marsden Road Public School. Marsden Road Public School is considered a ‘disadvantaged’ school, located in New South Wales. To really drive this point home:

“If we were sending 100 kids to high school, 90 of them couldn’t read or write as per their age — and that’s failing those students.”
- Manisha Gazula, Principal of Marsden Road Public School, as quoted by ABC News

There’s no doubt that these results were of significant concern; however, that’s not the case for Marsden Road anymore. In fact,

“Today Marsden Road Public School is a school transformed, with students achieving above the state average in writing and making above average progress in reading and numeracy.”
- Conor Duffy, ABC News

To really highlight these results, please see the following snapshot taken from the MySchool website for Marsden Road Public School:

What you will notice is that more than half of Marsden Road’s students make above average progress in all of reading, writing and numeracy, which is impressive in itself when you consider the fact that many students come from backgrounds of significant socioeconomic disadvantage.

What’s the reason for this change?

“Since adopting explicit instruction the school has transformed its NAPLAN results and sustained them, with life-changing results for students.”
- Conor Duffy, ABC News

Well. That’s telling.

There are countless more schools of varying backgrounds of socioeconomic advantage where you can find similar results, including secondary schools. I’m sure more examples will come out once 2026 NAPLAN results are released more publicly.

The point is this though. Whilst some critics are suggesting online that leaders chose a direction - mainly, explicit instruction or evidence-informed practice - committed to it, and are now blindly following it, it’s quite likely the opposite. For many school leaders, the results came first. They were the reason for the change.

Explicit instruction didn’t create the problems currently visible in our NAPLAN data. Persistent problems in the NAPLAN data are instead what drove schools to consider explicit instruction in the first place.

perhaps you should read the whole dataset

I agree with critics on this part. We should absolutely be reading the entire dataset. Please. Do it.

NAPLAN is valuable precisely because it gives us a common measure that allows us to see patterns we might otherwise explain away locally. For instance, NAPLAN consistently reveals a pattern of underachievement in many Australian students - a tragedy that isn’t shared evenly. As Anna Stokke said in her keynote address at researchED Melbourne on the weekend:

“All you can tell from a standardised test is a kid’s postal code”
- Anna Stokke

As is so often the case, it is the students who are already facing the greatest barriers to educational success who are most affected. To put this into context, in 2023 NAPLAN, Year 3 students from disadvantaged backgrounds were already over a year behind their advantaged peers in reading, with this difference increasing to more than 5 years by Year 97. This is exactly why reading the full dataset is important. What that dataset does is highlight just how unevenly educational underachievement is distributed.

The reality here is that a child from a more advantaged background has more opportunities to compensate for instructional deficits. As David Didau and Rikki Cole put it in their book, ‘Writing Fitness: From Beginning to 5K’8:

“More advantaged students will tend to be academically successful despite the choices we make, whereas the performance of less advantaged students is more likely to be because of what we’ve done”
- David Didau and Rikki Cole, Writing Fitness: From Beginning to 5K

The students who most need our instruction are also those who pay the highest price when that instruction is ineffective. This is exactly what our NAPLAN data reveals, and it is exactly what gets lost when persistent underachievement is weaponised against the very reforms designed to address it.

The fact of the matter is that we have the evidence of what is most likely to “work”. In fact, the principles underpinning explicit instruction are well-researched using a variety of methods - from randomised control trials through to meta-analyses, and even case studies of individual schools (like those mentioned above) who turned around their NAPLAN results. Choosing not to at least consider the way evidence can inform practice is instructional neglect, and the impacts of that aren’t shared evenly. Likewise, choosing to constantly cycle between different educational ‘fads’ doesn’t give any one technique a chance to make a difference. There’s a cost to constantly switching.

And realistically, as recited by Haili Hughes in her presentation over the weekend:

"If you’re not using evidence, you must be using prejudice”
- Sir Kevan Collins

So yes, please read the whole dataset.

Read the improvements and declines and the columns that make your preferred argument uncomfortable. But also read a little deeper.

Perhaps read into the socioeconomic gaps and significant proportion of Australian students who are not proficient at basic literacy and numeracy. Consider the years of persistent underachievement that preceded the reforms currently being blamed for failing to fix them quickly enough.

And then ask yourself which students can afford for us to get this wrong.

By all means, look for evidence that the new direction of Australian education might be wrong. But don’t become so determined to prove that it might fail that you forget to look at who was already being failed by the old one.

Share

“Ability is the consequence, not the cause, of what children learn.”
- Graham Nuthall, The Hidden Lives of Learners

It’s currently 6:30 am on the Sunshine Coast. Getting here involved a pre-4 am alarm and a drive in the dark - both decisions of mine that I’ve questioned at points today. With that said, it’s feeling far more worth it now. My dog, Pippa, has had her beach walk and swim, I've got a coffee, and I get to write this with a pretty spectacular view.

It’s also pretty good motivation to finally tackle this article, which has been sitting as a mildly provocative title-only in my drafts for weeks. Not because I didn't know what I thought, but because I knew exactly what I thought.

It starts with this.

Over the past few months, I've found myself coming back to the same phrase repeatedly:

Treat ability as an output, not an input.

What has surprised me is that a phrase I assumed was largely uncontroversial has turned out to be anything but. That in itself is concerning to me, because if, as educators, we don’t genuinely believe all of our students are capable of learning, then why even teach? At that point, instruction becomes little more than a sorting mechanism that rewards students who arrive to our classroom with greater prior knowledge - knowledge that is often the product of advantages well beyond what a student can control.

That’s perfectly reasonable if you see the role of an educator as identifying talent. It’s a little harder to defend if you believe our role is to develop it.

every inaction has a consequence

It is far easier to explain a student's struggle as the result of something that was predetermined before they ever walked into your classroom. Why? Because the alternative likely means admitting that our instruction wasn’t enough somewhere along the line.

That’s not to suggest that every student who experiences struggle in their schooling does so due to instructional neglect. Of course that’s not the case. Students do arrive to our classrooms with different experiences, prior knowledge or starting points - and that’s not to mention differences in working memory capacity or processing. However, by the same token, if our instructional design is underpinned by the assumption that those differences determine what students are capable of learning, then we're allowing our expectations to dictate our instruction, rather than allowing our instruction to shape what students become capable of.

I’m guilty of this. I’ve written on streaming/setting/grouping for instruction a couple of times previously (see here and here), and one of the concerns I consistently raise is that teachers of lower-prior attainment classes will often limit the curriculum that they expose students to in the first place. I say this because it’s so easy to do. Very early in my mathematics teaching I had a lower-prior attainment class, who I avoided using the word ‘coefficient’ with, because I’d decided that it would be too challenging to introduce subject-specific vocabulary to students who already struggled. You know what a better alternative would have been? To just teach them.

I had no right to decide that my students were incapable before I’d even given them the opportunity to learn something. That’s the danger of this belief.

The moment we decide what a student is capable of learning, we begin designing instruction around that assumption. We remove challenge and narrow the curriculum before they’ve even had the opportunity to engage with it, withholding the very knowledge that might have changed what they were capable of doing in the first place.

The students who need the great exposure to knowledge often end up receiving the least. And then, when they can’t do the thing we never truly taught them to do, we point to that outcome as evidence that our original judgment was right.

The tragedy is that we never get to know what those students might have been capable of. Not because they couldn't learn it, but because we decided not to teach it.

That brings me back to my original statement:

Treat ability as an output, not an input.

knowledge drives ability

I started this post with the following quote from Graham Nuthall.

I love this quote because it challenges one of the most common and damaging assumptions we make about learning, and does so so succinctly. More importantly, it shifts where responsibility sits. If ability is predetermined, instruction can only ever do so much. But if ability is, at least in part, a consequence of learning, it returns a sense of agency to both the learner and the educator. Instruction becomes one of the mechanisms through which ability is developed.

That, to me, is a far more empowering way to think about teaching. It’s a reminder that what we teach, how we teach it, and the opportunities we create genuinely influence what students become capable of doing. Equally, it means that a student’s past challenges don’t have to define their future success.

Don’t believe me? Based on the pushback I’ve received over the past few months, I suspect you’re not alone, so let’s unpack that a little.

To be a little controversial, I actually don’t think there are many - if any - people who flat out disagree with this logic. I think we apply it selectively.

Let me prove it to you.

We readily accept that training produces better athletes. We know not everyone will become an Olympian, but that doesn’t stop us believing that coaching and practice can dramatically improve performance. Likewise, we don’t believe learning an instrument is reserved for only for musical prodigies, and nobody else could possibly achieve any musical success. And up until a couple of years ago, my own cooking was so abysmal that a good friend once described it as "genuinely terrifying." Fast forward to today, and I love pulling off multi-course dinner parties, and was actually excited to receive a cast iron pan for my birthday last week. It’s wild the power of practice makes perfect1.

In all of those cases, ‘natural ability’ doesn’t change with practice, but it also doesn’t stop someone excelling with the right guidance. That’s the part we tend to overlook in education. In other contexts, coaching, practice and training are spoken about as being ways to develop a skill, but in reality, their value lies in what they produce. Knowledge.

Seriously though. Knowledge isn’t just some arbitrary thing we accumulate. It fundamentally changes what we’re capable of understanding or accessing.

Every new piece of acquired knowledge becomes a foundation for future learning. The more we know about a topic, the easier it becomes to learn more about it. You’ve almost certainly experienced this yourself. Think about how much easier it is to learn new information about something that you already know a lot about. It even feels as though the new knowledge is just slotting right into place.

The fact that knowledge builds on knowledge is precisely why it’s so powerful. John Sweller captures this exact idea when discussing individual differences that exist in learners:

“While there are many factors that lead to individual differences, the acquisition of knowledge is a major, possibly the only cognitive factor that is readily malleable and alterable.” - John Sweller

Now THAT is a powerful statement.

Because that’s not a claim that individual differences don’t exist. In fact, that’s an explicit acknowledgment that they do, with a full recommendation built in. His point is simply that, of all the factors contributing to those differences, knowledge is one of the few we can meaningfully influence. That’s remarkable news for educators because it means our role isn't simply to work around students' current ‘ability’, but rather to build the knowledge that changes it.

Apart from knowing that we have the ability to influence knowledge, I think it’s worth noting as well how knowledge can ‘compensate’ for some other individual differences that exist in learners. When I presented on this topic recently, I spoke to some of the differences we perceive in learners and how increasing knowledge can either get around them, or at least mitigate their impact.

For instance, take learners with different working memory capacities for example. Memorising and recalling a sequence of 10 random letters, such as ‘KWPSLFMQFI’, is likely going to be far harder than memorising ‘MEMORISING’. Obviously. Because you know the word ‘memorising’. It’s a piece of knowledge you already have. And that piece of knowledge is the reason you can process that sequence of 10-letters as one chunk, rather than ten.

That’s exactly why prior knowledge matters so much. It doesn’t just help students to learn new content, but it changes how demanding related learning is in the first place by allowing learners to chunk items, freeing up working memory. That in itself is a case for building up prior knowledge for learners with differing working memory capacities.

Similar can be argued for a lot of differences in individual learners. If they have the right prior knowledge, learning is inherently easier, and of course, knowledge is something we can directly influence. That’s why I find it so difficult to accept the idea that ability should be treated as an input into teaching. It’s not to say that individual differences don’t exist, but rather that they don’t discredit the importance of knowledge.

yeah but nobody is arguing that knowledge is bad

Maybe not out loud, but it’s implicit every time a lesson plan excludes content because it’s “too difficult” or will take “too long” to teach the learner. Rather than expose students to the knowledge they need, is it really a better case to just cut it out of the curriculum? That’s not protecting students. That’s protecting your own expectations.

Of course, none of this is to pretend the practical reality is simple.

I’m a practising classroom teacher, so I’d be doing everyone a disservice if I suggested that building years of missing prior knowledge into a set curriculum is an easy thing to do. It isn’t. There are many days that I’ve spent several hours planning how I can fill in gaps from previous years whilst still meeting core content as efficiently as possible.

In my daily work as a teacher and working externally with schools, there’s the constant decisions of whether to slow down and build the knowledge students need or to keep moving through the core content, as required. I don’t have a perfect solution to that, although I do have some practical strategies that I’ll likely talk to in a future post (access questions2 + retrieval).

What I do know though is that removing some knowledge from the curriculum because we assume students won’t be able to access it certainly isn’t the answer. Likewise, not building in opportunities to fill in some of those previous gaps they may have formed is almost equally not the answer. If knowledge changes what students are capable of doing, then denying students access to that knowledge simply guarantees they’ll remain less capable than they might otherwise have become.

and if you don’t believe that?

I recently said at a conference that I would have serious questions for any educator who applies the blanket logic that students have a fixed ability and aren’t capable of learning beyond that. Whilst harsh, I standby that statement wholeheartedly. If we do honestly believe that there is little we can do to change what a student is capable of understanding, why would we bother putting in that extra effort to meet them where they are and support them to develop from there? Quite frankly, I don’t think we would. Teaching only makes sense if we believe instruction changes what students know.

And if teaching changes what students know, it also changes what students become capable of doing. That, in part, is why I keep going back to that same phrase:

Treat ability as an output, not an input.

You need to believe that your instruction can actually make a difference, otherwise the most dangerous belief you’ll never write into a lesson isn’t that some students will struggle. It’s that you’ve decided which students always will.

Share

Because it’s me, I also proceeded to tempt fate once more. Less than 24 hours after my symptoms had finally disappeared, I found myself camping last weekend. You read that correctly. Camping. Like sleeping outside. This might not sound like a big deal, but if you know me even in the slightest, you will understand why that news was met with a significant amount of disbelief from pretty much everyone who has spent more than 5 minutes with me. For those of you who haven’t, I can summarise my usual sentiments towards camping with the following response I threw out the last time someone asked if I wanted to go camping:

“You know you can choose to just not do that, right?”

However, in some bizarre twist of fate, I enjoyed my time sleeping under the stars, so much so I’m considering doing it again.

Why am I telling you this? Certainly in part because this is a little off-brand for me, but mainly to talk about why it was so successful. See, if you asked me why I’ve previously been so reluctant to camp, I could probably summarise it in a few key points:

• I hate the cold1

• I don’t know what I’m doing - it’s been a good many years since I’ve been camping and, frankly, I’m scared.

So what made this different? Well, I was in Queensland and it was over 20 degrees (68 Fahrenheit if that’s your jam) during the day and, whilst the nights were cold, they weren’t Melbourne cold2.

More importantly, I'd describe this as a gentle on-ramp to camping. There was a house with actual plumbing. I didn't have to set anything up myself, and there was enough moral support around that I never had to make decisions about things I know absolutely nothing about. It was, essentially, camping with scaffolds.

“But, Caiti, that isn’t REAL camping!!!”, I hear you say.

I’m not going to argue that. I’m under no illusions that I really took the scaffolded option here. However, that scaffolded option is what got me to do it in the first place. Those scaffolds made me feel at ease about something that I would have previously described as being remarkably unappealing, and they gave me a chance to learn some elements myself. They’re the reason I want to do it again, and likely the reason that over time I’ll gradually delve into slightly-more-real-but-also-not-totally-serious camping in the future3.

You know what wouldn’t have helped that? Either telling me not to bother since what I did isn’t considered “real camping”4 or pandering to my personal wishes and letting me sleep in a house. Neither option would have moved me any closer to becoming someone who enjoys - or is even capable of - camping. I hope you can see where I'm going with this.

We do the same thing in education. Sometimes we throw students into the deep end in the name of independence and watch them flounder. Other times, convinced they’ll never be able to access the core curriculum, we lower the bar altogether and replace meaningful learning with tasks designed simply to keep them occupied. Both approaches teach the same lesson: this isn’t for you.

There’s a common criticism that explicit instruction creates learned helplessness by providing too much support5. But if learned helplessness is the belief that your efforts won’t lead to success, we should ask a different question.

Who’s really teaching helplessness?

you’ve got it backwards

Let’s start at the beginning. What’s the wild claim that this week’s epic takedown will centre around?

Explicit instruction causes learned helplessness.

Firstly, to make sure we’re clear on what we’re talking about, I’m going to use the following definition of learned helplessness:

“In psychology, learned helplessness is a state that occurs after a person has repeatedly experienced a stressful situation. They believe they cannot control or change the situation, so they do not try, even when opportunities for change are available.” - What is learned helplessness?, Medical News Today

Essentially, this is the argument that the scaffolding, supports, highly-guided instructional design that constitutes explicit instruction creates learners dependent on teaching often6 likened to “hand holding”. Of course, that is hardly reminiscent of the real world where nobody is sat beside you at every waking moment helping you to solve problems or break an essay into individual sentences. Gee, it’s not even reminiscent of an exam, where students are usually on their own.

However, by arguing that explicit instruction causes learned helplessness, does that not suggest that alternative methods are more likely to produce independent learners?

Because I would argue that they are far more likely to teach learned helplessness.

just throw ‘em in the pool

Alternative #1: I learned to swim by being thrown in the pool - they’ll either drown or learn to swim pretty quickly.

If you’re lucky, they might also escape the above scenario without too much trauma too. What a win.

This is the idea that if students need to know how to ‘problem solve’ by themselves, you should start them there. Begin how you want to end, or something like that. I’m not even saying just giving them problems and not teaching. It could be modelling a full problem in front of them, then sending them off to work independently.

I’m also not saying this won’t work, because it will and has done so for some individuals. There was a time in my schooling where I was using the internet to teach myself maths in the absence of a teacher willing to do it, and I turned out fine. However, it would be misplaced to assume that my personal experience can be generalised to an entire population. As I often say, the students we neglect by not providing access to the core content through our instruction are often the ones who need the most support. They’re the ones who are starting from a lower prior knowledge point and more often than not, those from disadvantaged backgrounds in some form are over-represented here. I was lucky enough to have an incredible older brother who I could ask maths questions to. Not every young person has that, just like not every family has the means to pay for a private tutor to fill in the gaps left by instruction.

Some students will certainly succeed despite this. But if your instructional approach relies on students already possessing the knowledge, skills or support to teach themselves, then it will inevitably leave many others behind.

And what happens to those students? They don’t simply fail to learn the content. They begin learning something else.

Every time a student is presented with learning activities that they don’t yet have the knowledge to solve that belief that effort doesn’t lead to success is reinforced. Why even bother trying? Experience has taught them that trying doesn’t change the outcome.

Wait a moment. That sounds familiar. What was that definition of learned helplessness again?

“In psychology, learned helplessness is a state that occurs after a person has repeatedly experienced a stressful situation. They believe they cannot control or change the situation, so they do not try, even when opportunities for change are available.” - What is learned helplessness?, Medical News Today

Who would have thought attempted drowning was an ineffective way to teach people to swim? Everyone. Literally everyone would have thought that.

if they can’t swim, just let them play in the shallows

Alternative #2: Not everyone is cut out for swimming. Just let them have a play where they can’t get hurt.

If you don’t try, you can’t fail.

To be clear, I’m saying that to the teachers, not the students. When you know the lesson could be challenging, it' feels far kinder to just not teach it. This alternative is what happens is students find a subject challenging - likely due to lower prior knowledge, or maybe something else. In my experience, this is usually the appropriate taken to teaching ‘foundation’ or ‘low prior-attaining’ classes in streamed/grouped models, but can sometimes occur when some students in a class are given an alternative activity to their classmates7.

Teachers recognise the frustration that comes from years of feeling unsuccessful, and they do what they genuinely believe is the kind thing: they replace the challenging learning with a curriculum-adjacent task that feels more accessible, more engaging, and more achievable. The students are happy in the moment. But the attainment gap between these students and their peers increases.

In trying to spare them the feeling of failure today, we quietly reduce the likelihood of success tomorrow.

Every time a teacher makes the decision that the core curriculum is ‘too much’ for a student or group of students, we make a decision about their future capabilities. And for each year that goes by where teachers give the lower prior-attaining students easier tasks or vaguely curriculum-adjacent activities, it becomes more difficult for these students to catch up. They weren’t given the chance to make that decision for themselves. It was made for them. In other words that I’ve likely used before, we give the students who need the most support the least.

The real tragedy is that this can become self-fulfilling. The less exposure students have to the core curriculum, the fewer opportunities they have to build the knowledge that would make future success possible. As the knowledge gap between these students and their peers widens, the original judgment about these students appears increasingly justified.

This is two-fold though. Students also notice. They notice that while some classmates are grappling with the real work, they’re completing something different, which, at times, can be frankly insulting8. They notice that they are rarely - if ever - given the opportunity to even attempt the same work, because the decision has already been made that they won’t succeed. And to make matters worse, in some cases, they still sit the same assessment task or exam as their peers who are actually being taught the core content. Wow, so kind.

Over time, it’s pretty easy to why these students will begin to say that school isn’t for them or they’re “not a maths person” (or any other subject equivalent). It’s easy to see why they don’t try - partly because they’ve always been told they shouldn’t, or at the very least, don’t have to. In protecting students from today’s challenges, we unintentionally teach them that tomorrow’s challenge isn’t even worth attempting. Now that too sounds familiar…

“In psychology, learned helplessness is a state that occurs after a person has repeatedly experienced a stressful situation. They believe they cannot control or change the situation, so they do not try, even when opportunities for change are available.” - What is learned helplessness?, Medical News Today

Oh yes, that sounds an awful lot like learned helplessness.

floaties aren’t forever

I’m not loving the pool analogies, but we’re in too deep to stop now. See what I did there?

Let’s flip to the other end of the scale for a moment and talk about highly guided instruction, or ‘over-scaffolding’ if you prefer.

I’ve explained how it is two different methods of providing minimally guided (or not-at-all guided) instruction can generate learned helplessness, but surely too much guidance is just as problematic, right? WRONG. Kinda…

Since Nidhi Sachdeva and Jim Hewitt have written a terrific post already explaining why explicit instruction doesn’t cause learned helplessness here, I’m not going to go too deep into this, but I do want to briefly talk to the argument that highly guided explicit instruction causes learned helplessness.

I’m not even going to deny that when I’m teaching novices new content, I do scaffold well over what would be required in the real world. You know what, have an example from one of my lessons this week:

“We’re finding the circumference of a circle. What is the circumference?”

…

“Yes, it’s the distance all the way around the edge of a circle. When we find circumference, we use the structure FSA. What does F stand for?”

…

“F stands for formula. Show me the formula for the circumference of a circle.”

…

“Great, it’s C = 2*pi*r. We’ve done F, so we’re now up to S. What does S stand for?”

…

“Perfect, it stands for substitute. That means we substitute in values in place of our pronumerals. What pronumeral are we going to substitute a value in for?”

…

“Yes, r. We’re trying to find C, so we leave that as C. 2 is a number and pi is a number. And what does r stand for?”

…

“Yep, radius. Now substitute in the radius.”

… and so on….9

Yep, over-scaffolded. Nobody is structuring an exam like that. Cancel me, I guess.

The mistake is assuming that’s where the instruction remains. It’s not. It’s a start point. Every additional prompt is introduced with the intention to be removed - as quickly as I can do so while still supporting my students to experience success.

Over the course of that teaching sequence above, I gradually faded my guidance, using frequent checks for understanding to decide when students were ready for the next step. Eventually, all that remained on the screen was a single instruction:

Find the circumference.

No prompts. No scaffolding. No annoying hints from me. Just the question.

Would you be surprised to hear that not only were my students successful, but the structure of their responses was perfect?

Well-executed highly guided explicit instruction doesn't result in student success because the teacher is doing the thinking for the students, or ‘hand holding’, if you will. The success comes from the fact that barriers preventing access to the core knowledge are temporarily removed, then gradually reintroduced as the students gain the knowledge to overcome them independently.

Within this, I’ve become increasingly fond of what I’ve come to call access questions. Before introducing new content, I deliberately identify the prerequisite knowledge students will need simply to access the major learnings for the lesson. Rather than teaching the content first, waiting for students to fail, and then patching up the gaps during my regular instruction, I check - and, where necessary, teach - that prerequisite knowledge upfront during my warm-up. For some students, this is just nice practice of the knowledge they’ll require in that lesson, but for others, it’s the thing that makes the lesson accessible.

Now, if I were to go back to that definition of learned helplessness from earlier, I’d say that’s actually the opposite of learned helplessness:

“In psychology, learned helplessness is a state that occurs after a person has repeatedly experienced a stressful situation. They believe they cannot control or change the situation, so they do not try, even when opportunities for change are available.” - What is learned helplessness?, Medical News Today

As someone who has taught students across the full spectrum of prior attainment in mathematics - including those who have reached their final year of high school convinced they’ll never be good at maths - I can easily say that the guidance and scaffolds that allow students to experience even small doses of success that they might not have ever felt before are a great way to show them that they can change this situation. Not only does it make success possible, but it makes future success possible.

That’s hardly learned helplessness. In fact, I’d go as far as to call that learning.

so who’s really teaching helplessness then?

Learned helplessness isn’t created by supporting students to learn. A far more effective way of creating learned helplessness would be to have students repeatedly experience failure, or quietly exclude them from the very knowledge that would allow them to succeed in the future.

If we want students to become independent learners, we don’t get there by withholding the instruction they need. We get there by making success possible often enough that they begin to believe their own effort can influence the outcome, then gradually fading that support as their knowledge and confidence grow.

Now, it would be wrong of me not to acknowledge that faded guidance is a critical component of explicit instruction. The goal was never for my students to need my prompts forever, just as the goal isn’t for someone learning to swim to wear floaties forever.

And, to bring it back to camping, nobody made me enjoy camping by throwing me into the wilderness with a sheet of tarp and some poles, and wishing me luck. Equally, nobody convinced me I was capable by letting me stay inside the house all weekend. I’ve become more willing to camp because I was given enough support that I could actually experience success.

Weirdly, this is remarkably similar to how this works in the classroom.

Independence isn’t the absence of instruction. Independence is what remains after good instruction.

Share

5

Amazing post here by Nidhi Sachdeva and Jim Hewitt countering this:

The Science of Learning

No, Explicit Instruction Does Not Cause Learned Helplessness

Read more

3 months ago · 106 likes · 31 comments · Nidhi Sachdeva and Jim Hewitt

Did I really need to write another1 post on autonomy and/or low-variance curriculum and/or consistency in teaching? Possibly not. But, by that logic, we could ask if I ever really need to write at all, and I seem determined to keep doing that, so here we are.

What has prompted yet another post on this topic that I should possibly be bored of? Why, edu-social-media of course! The topic of low-variance curriculum is #topical, as the kids would say, and that means I write about it.

This time around though, I wanted to work to dispel a few more myths, worries and urban legends surrounding the terrifying low-variance curriculum. After all, most monsters are far less frightening once you turn on the lights.

If you want my first set of myths of low-variance, see this post how to take down an argument and not an individual: low-variance doesn't mean no variance.

what, why and all of those important questions?

“The word low-variance is tossed around a lot these days…” - Homer Simpson if he was an active participant in edu-media at present, I imagine, probably

Despite its frequent appearances, I’m also reasonably convinced we’re not necessarily clear on the definition of low-variance, so let’s start there.

Beginning with ‘variance’ itself, I quite like the following description taken from Rob Marchetto based on a webinar by Reid Smith for Think Forward Educators:

“When we think of curriculum variance we are talking about the variance in taught and learnt content as well as variance in instruction. We need to think of a range, from a continuum of lots of variance (complete variance) to one of next to no variance.” - Rob Marchetto

Essentially, variance in our curriculum refers to the differences between classes in the content - both what is taught and what is learnt - and variance in how it is taught. For simplicity, let’s also assume that we’re referring to classes of the same subject and/or year level - essentially, those with the same learning goals.

This could be all Year 3 classes in a school, for instance, all Year 8 Mathematics classes, or all Year 12 Economics classes. Basically, they’re the classes doing the same stuff.

As above, variance exists on a continuum.

“High variance”, shown on the left below, is something I suspect is reasonably uncommon. Without going too far down a tangent on what high variance could be, it would consist of significant differences in what is taught, what students learn and how it is taught. In its most extreme form, this could be two classes studying the same subject, but entirely different content. Perhaps, one Ancient History class studying Julius Caesar, whilst another studies Augustus Caesar2, with no intentionally shared methods, learning outcomes or instructional methods. Imagine the wild, wild west of teaching.

I would imagine (hope) most curriculum sits a little further to the right than that.

Let’s move to the other end of the continuum: no variance. In this scenario, every class is taught the same content, in the same sequence, and formally, there should be no differences. None. Probably also not possible.

The closest thing to this would likely be a Direct Instruction (“capital D, capital I”3) program, or scripted lessons. I’m not going to discuss those here, but will acknowledge that they sit more to the right end of the continuum if we’re talking about a low-variance curriculum. At the same time, I will also add in the brief caveat that I am for Direct Instruction (“capital D, capital I”) programs; however, they’re rarely what people are talking about4 when speaking about low-variance curriculum. Let’s just take our fear of scripted lessons and put that to the side, because that’s not our conversation right now.

Instead, we will focus on low-variance, which still sits more towards the ‘no variance’ end of that continuum than the other end, but probably isn’t quite as daunting as it sounds.

I say that as someone who has been very open about having taught in a school that would have some the lowest variance curriculum in Australia, and has also said I was very worried about my loss of autonomy before I started there. Genuinely, I didn’t like the idea - until I did, which only really took me a couple of weeks of getting my head around a new school, and now I would love to low-variance my way through life. Total convert.

What constitutes low-variance?

It’s a continuum, so the definitions vary a little, but usually it consists of the following:

• Shared learning outcomes/goals

• Shared curriculum materials

• Agreed instructional methods

Wait, that doesn’t sound so bad?

Yes, and I would agree. Shared learning goals, common curriculum resources and agreed instructional approaches do seem fairly uncontroversial. After all, most schools/curriculum bodies already mandate the first of those already as part of whatever state/national curriculum they have.

Yet edu-media loves to frame low-variance curriculum as the sworn enemy of teacher autonomy. Why? Because saying you’ll lose your autonomy is scary. Ah yes, fear-mongering again. You can always tell an argument is weak when it relies on over-exaggeration and fear-mongering to recruit followers.

On that note, let’s break down some myths.

Subscribe now

myth 1: suppression of professional expertise

Not all variation is valuable variation.

One of my favourite myths around the use of a low-variance curriculum is that is suppresses professional expertise. I read an article recently that referred to the kind of evidence-informed practice underpinning a low-variance curriculum as a “dumbing down of pedagogy”5, reinforcing the view that teachers’ expertise is somehow rendered obsolete under such a model, almost as if PowerPoint might step in and do the teaching for you.

Often times, this argument is coupled with a comparison to another profession. Cue the classic: but doctors exercise professional judgment all the time.

Yeah. Informed by the evidence. That’s a pretty central part to professional judgment. But what would I know?

The more interesting question is probably whether or not professional judgment requires high variance to exist. After all, doctors - since we love that comparison in edu-media - aren’t out here individually deciding on their own infection control protocols or prescribing non-approved treatments based on personal preference. Their expertise is exercised within a framework of agreed best practice, informed by evidence and refined collectively over time.

Nobody is interpreting this as a suppression of professional expertise.

Yet in education, there is a concern that a low-variance model completely stifles any need or use for professional expertise - a concern that is, frankly, insulting.

If you’ve read about my 100+ slides per lesson, you will likely know that on these, I exercise a lot of professional judgment. Each time I run a whole class check for understanding on the humble mini whiteboard6, I need to quickly assess any misconceptions/misunderstandings and what my next course of action is - all of which I do using my professional expertise. Quite likely more than 100 times in an hour-long lesson. And that’s not to mention the freedom to use your professional expertise to tailor your delivery, ramping up the energy and whiteboard banging when your class is half-asleep, or slowing down and dropping your volume when it’s the last hour of the day after a lunchtime house event and your students are elevated.

That doesn’t even begin to touch on the decision to put up some questions or give your students a quick task so you can step outside for a moment to support a student who isn’t having a great day. Likewise, it doesn’t stop you putting a pause on the lesson content because you’ve realised you’re teaching integral calculus to students who can’t divide fractions, and you’re going to need to do a brief detour into that.

In fact, the main difference is that under a low-variance model, I am no longer repeatedly exercising my professional judgement over decisions about what examples to teach or how I can possibly plan an effective, well-sequenced lesson at 11pm the night before. Instead, the collective decisions and shared resources free me up to direct my professional judgments towards the infinitely more complex work of responding to the students in front of me. In the same way that teachers in a team might share an activity or worksheet or set of questions, the bulk of the professional judgment teachers exercise in the classroom is in noticing what’s happening in that moment with the young people in front of them, and knowing how to respond - a response that isn’t dictated by a set of slides.

I said recently, you’re not handcuffed to the slides. In fact, in a recent episode of When Will They Learn? Podcast with Greg Ashman, I even asked Greg what he would do if he realised his students were missing a skill they needed to even be able to access lesson content - a skill not present in the shared curriculum slides. Unsurprisingly, Greg said he would stop and teach it, and even add some questions into the resource for next year to save other teachers having to do the same on the spot. Some might say, he would exercise his professional judgment to make a decision about what his class needs in that moment.

Low-variance curriculum doesn’t eliminate professional expertise. It simply asks us to consider whether that expertise is best spent reinventing Year 8 fractions for the tenth year in a row, or whether it is better directed towards the profoundly human work of teaching itself.

Professional judgement does not disappear within a framework of agreed best practice, informed by evidence. If anything, that framework exists because of it. And, to put it bluntly, if other professions can still exercise their knowledge, expertise and professional judgment under agreed frameworks, I think teachers will cope just fine.

myth 2: won’t someone think of the autonomy?

But without autonomy to be the coolest, how can I leave a legacy as the rockstar teacher to all my students?

Better question: why would you want to? Basing your self-worth off how cool children think you are seems like a bad idea, if not kinda really weird and a bit sad.

For real though, one of the more passionately charged criticisms of low-variance curriculum is that it removes teacher autonomy. After all, teaching is a profession, and professionals often value the ability to exercise choice in their work. However, the issue is when that autonomy comes at the cost of student learning.

It would be wrong for me to flat out say that a low-variance curriculum doesn’t have some impact on autonomy; however, it would be equally as wrong for me to say that having a job in the first place doesn’t impact my autonomy. I love my job, but I also really love spending days at the beach with my dog, and given the autonomy to do whatever I want, I suspect beach days with my dog might win a few more times than going to work. In every profession, there is an expectation that employees abide by particular rules and responsibilities. Inherently, that removes some kind of autonomy. No teacher is - nor should be - completely autonomous.

Where this becomes more nuanced is that not all forms of autonomy are equally beneficial, nor are they equally important.

The concern is often framed as though low-variance curriculum removes teacher autonomy altogether. In reality, what it tends to do is redistribute it. Teachers may have less autonomy over decisions that have been made collectively - such as learning goals, curriculum sequencing, shared resources or agreed instructional approaches - but they retain autonomy in the countless decisions they make in the act of teaching itself. Similar to the arguments around professional expertise, it is possible to make decisions or put your own flair on a low-variance lesson.

Goodness, I can tell you that nowhere within our agreed instructional methods on our shared low-variance resources for teaching decimal multiplication did it say that the decimal points should explode and that it should be taught by loudly exclaiming this while hitting the whiteboard - but hey, autonomy! Likewise, at no point did I need to tell my students the story about how I can never successfully order peppermint slice. Yet somehow, that too found its way into the lesson. I also can't recall any low-variance curriculum document stipulating that:

“If it’s important, it should be in pink. And if it’s REALLY important, it should be in glitter” - Caiti Wade, frequently when teaching maths to teenage boys.

And yet, some of my students could probably recite that back to me years later.

Believe it or not, it’s possible to show off some personality and autonomy under a low-variance model. Shared curriculum resources or instructional methods don’t erase teacher personality, humour, warmth or responsiveness. Instead, they establish a common foundation for the quality of education that students receive.

Of course a low-variance curriculum somewhat impacts teacher autonomy. Working a paid job as a professional tends to do that. But what level of autonomy is realistic or even desirable?

Because if the price of absolute autonomy is that students receive vastly different educational experiences depending on the luck of the timetable, then perhaps some constraints are worth embracing. Perhaps students deserve consistency in the things that matter most, like the instruction and curriculum, while teachers retain autonomy in how they bring that learning to life.

After all, students don’t need thirty different versions of decimal multiplication depending on who is teaching them.

But they probably won’t mind thirty different stories about peppermint slice.

Autonomy isn’t just the absence of any boundaries. Autonomy is what you can still do within them. And only some of those things are reasonable, and worth doing.

myth 3: one size doesn’t fit all

I think you’ll find that’s mostly just a catchy phrase and doesn’t apply all that well to how our brains learn.

Let me remind you:

“Children are more alike than different in terms of how they think and learn.” - Daniel Willingham, Why Students Don’t Like School

Of course, this doesn’t mean that all children are identical. Students differ in their prior knowledge7, interests, motivations, confidence, and the support they require. Responsive teaching remains an essential part of the profession.

However, recognising these differences is not the same as concluding that every student requires an entirely different curriculum or fundamentally different instructional approaches. In fact, I would highly doubt anyone using the “one size doesn’t fit all” argument is teaching thirty different versions of whatever question to their class. How does a teacher stood in front of an entire class tailor their explanations so that they are different for each individual simultaneously?

They’re not.

Teachers advocating for this would agree that whilst it would be ideal to individually tutor each student, that is unfortunately not a practical reality in a modern classroom where you’re responsible for 20-30 young people at once. The alternative is to teach the method that is deemed likely to work for most students and address any individuals as needed.

I hate to break it to you - I actually don’t; I get a lot of joy out of this - but that’s exactly what a low-variance curriculum does.

Seriously, it’s about using the evidence to inform instructional decisions around resources, examples, methods, explanations and sequencing, with the aim being to teach in such a way that it’s most likely to be effective for most students most of the time. With those it doesn’t quite land with, individual interventions come into play, just in the same way they always would.

The reality is that the broad principles underpinning effective instruction are remarkably consistent. For example, novice learners benefit more from guided, explicit instruction than from being expected to independently discover concepts through minimally guided problem-solving. This isn’t because students are identical. It’s because the cognitive architecture through which humans learn is pretty much the same for all of us.

Any good low-variance curriculum will have frequent checks for understanding built in to allow the teacher to respond to their class in the moment. We know that the knowledge students already have can change how or how quickly they learn new information. That is,

“That differential knowledge results in individual differences in expertise and provides a major source, possibly the only source, of modifiable cognitive differences between learners.” - John Sweller, 2024

The benefit of frequent checks for understanding is that it allows teachers to be responsive to the students in their room in that moment, and to know who requires individual interventions.

If I were to flashback to when I started teaching, where I would talk maths at my class for 20 minutes while they silently copied, I was still teaching a ‘one size fits all’ model, but the difference was that I was only picking up on individual - and collective - misconceptions when I then ran around the classroom as students worked, trying to help individuals, asses where they were at and manage the classroom. It was a remarkably inefficient and ineffective approach that’s taught a little too commonly in university teaching degrees. The difference now is that I know in the moment who doesn’t understand something, and whether that’s a whole-class issue (or close to it) that I can fix with collective correction, or whether it’s something I need to address with an individual intervention.

Acknowledging that students are unique does not require us to abandon the idea of shared, evidence-informed approaches to curriculum and instruction. I agree that individual students deserve responsive teaching. That just doesn't necessarily require individual curricula to get there.

pink, glitter & other educational crises

Low-variance curriculum is not a panacea. It will not solve every challenge facing education. Implemented poorly, it can absolutely become rigid, performative, or compliance-driven. 100%.

But many of the criticisms levelled against it are not criticisms of low-variance itself. Rather, they are criticisms of extreme interpretations, poor implementation, or assumptions we’ve accepted without scrutiny. Did somebody say scripted lessons?8

The truth is probably less dramatic than would make for a worthwhile edu-media scare campaign. Teachers can exercise professional expertise within shared frameworks. Autonomy can coexist with consistency. Students can receive responsive teaching without requiring entirely different curricula.

In fact, I think many of the criticisms levelled at low-variance approaches are based on myths – persistent, well-intentioned myths, but myths nonetheless. And until we challenge them, we’ll continue framing the conversation as a choice between two things that may not actually be in conflict.

If you’ve made it this far, you may have noticed that the good old low-variance curriculum did not, in fact, prevent decimal points from exploding, peppermint slice stories from derailing lessons, or glitter from finding its way into secondary mathematics instruction.

It also didn’t eliminate professional expertise, completely destroy teacher autonomy, or require every student to receive precisely the same educational experience through some insidious educational dystopia in which every classroom is indistinguishable from the next.

What it may do, however, is ensure that students have access to a more consistent, evidence-informed education, regardless of whose classroom they walk into. And frankly, that seems like a goal worth pursuing.

Maybe the monster wasn't low-variance itself, but the stories we've told about it.9

Share

On Monday, I was lucky enough to attend a professional development session on responsive teaching and retrieval practice presented by Carl Hendrick himself. When I say I barely slept Sunday night because I was excited, I mean it1.

After an incredible day of learning that I thought couldn’t get better, I then got home to a butter bell delivery. Yes, a butter bell. Stay with me.

For those not familiar, a butter bell is a ceramic container designed to keep butter soft at room temperature using a water seal. Acquiring one has become an increasing personal priority since moving to subtropical Brisbane and discovering that, for roughly 80% of the year, my butter options are either hardened fridge butter that damages toast, or butter that you could consume with a straw. You might wonder why it’s taken me so long to invest in a butter bell2, and that’s because I’m what my brother describes as an ‘optimiser’, so obsessed with finding the perfect option that I delay major life decisions, such as butter bell purchases, for years.

Anyway, I finally found one online that I actually liked (read: it matched my general aesthetic), but of course they didn’t ship to Australia. At this point, I remembered a friend and former colleague from Victoria is a potter3, and so I sent him a picture and asked if he could make me one just like it. His response was that what I had thought was an aesthetically pleasing butter bell wouldn’t actually form the crucial water seal, and so was essentially just a dish with butter in it.

Because, while I eat butter regularly and had very confident opinions about butter bell aesthetics, what I did not have was any real knowledge about how butter bells function.

He still made me a butter bell of course. Beautiful, custom and far better than what I could have asked for - and he had designed it so it actually functions as a butter bell.

Somewhere between admiring my aesthetically pleasing spreadable butter and reflecting on the fact that I had spent a year trying to optimise a butter bell I did not remotely understand, I realised something: I had been remarkably opinionated about butter bells for someone whose expertise began and ended at eating butter.

And when you couple that thought with a day of listening to the incredible Carl Hendrick speak about how humans aren’t always great at recognising what good learning looks like, it got me thinking:

Everyone has an opinion on school - or rather, what schooling should look like.

Of course they do. A lot of people have been to school themselves.

However, attending school does little more to qualify someone to understand how learning happens than eating butter qualifies me to design a butter bell.

“but this worked for me”

And I love that for you, but…

Imagine I told you that I did well in maths because I always ate hazelnut chocolate right before my exam. That I swear by the mathematical benefits of it.

You would, I hope, have questions. Perhaps about causality. Perhaps about whether I’d considered all the maths study I’d done. Maybe about which brand I’d recommend. More than likely though, you would probably realise something especially obvious:

One person having an experience isn’t particularly compelling evidence that the thing caused the outcome.

Yet, conversations with people outside of education are filled with arguments only marginally less absurd than this. Seriously. Who needs a randomised controlled trial with isolated variables to determine causation, when we can have one person do a thing one time and base all future judgments off that?

Because a lot of people did go to school, it’s easy to fall back on the line, “but X worked for me” or “I was taught like Y and I can still read/write/insert other skill”. And to be clear, this isn’t to say an individual student couldn’t have been taught in a particular way and succeeded. Of course that happens. I can assure you that my older brother wasn’t well-versed in cognitive science when he taught me to read as a child4 and yet I seem to be doing okay in that space.

However, individual success stories are not the same thing as evidence that an approach is broadly effective. The fact that something worked for one person does not necessarily mean it is likely to work for all - or even most - students.

Spurred on by the critics of evidence-informed practice who love to use anecdotes about doctors relying on autonomous judgments over evidence, I raised this with a doctor recently to find out what really happens in the medical profession. The response I got was that just because a medication or intervention works on an individual doesn’t mean that it’s better for the population - and it might not even work for the same individual the second time around. That was also followed up with a statement on why randomised controlled trials are far better at determining medical interventions that are most likely to work for a broader population.

Isn’t this exactly what evidence-informed educators have been saying this whole time?? In fact, don’t I have a whole post on this EXACT idea here? And also another one here??

Like, I don’t want to be that person, but…

“All cognitive load theory effects are based on multiple, replicated, randomised, controlled trials.” - Sweller, 2024

Respectfully, kinda awkward that critics of evidence-informed instruction are out here trying to tell us doctors are also into guided discovery medicine over evidence. Especially when they’re literally not though.

Okay. I’ll stop5.

But it does summarise this point nicely. Sometimes an instructional technique will work for an individual; however, that is not sufficient to extrapolate those results to an entire population. Evidence-informed practice is about probability. To quote one of my previous posts, it’s about choosing what’s most likely to work for most students, most of the time.

looks can deceive

If it looks like learning and sounds like learning… it might not be learning.

Unfortunately though, there’s an added complication to this situation. Often, the things that feel or look like learning, might not be that. In his book, Instructional Illusions, Carl Hendrick speaks about how students can be emotionally engaged without being cognitively engaged, referencing that only the latter leads to learning.

This means that even when using evidence-informed practice, sometimes it doesn’t look or feel like what we think learning is supposed to, and that can make it a hard sell for students, their families, educators, and literally anyone who has tried to learn anything remotely academic6.

So, on that note, let’s talk about some of the classic tricks that we fall for: things that look or feel like learning, but often aren’t.

easiness trick

As we all know, every worthwhile human achievement should feel effortless.

In his well-titled book, Why Students Don’t Like School, Daniel Willingham explains that the human brain isn’t designed for the kind of academic thinking that is often done at schools - in fact, it actively tries to avoid it.

“People's minds are not especially well suited to thinking; thinking is slow, effortful, and uncertain. For this reason, deliberate thinking does not guide people's behavior in most situations.” - Daniel Willingham, Why Students Don’t Like School

One very effective way of avoiding thinking? Getting someone else to do the thinking for you.

By over-scaffolding or over-prompting, it is entirely possible to turn challenging learning into an effortless… experience.

Think: starts with E, ends in Xperience. Very effective as a prompt to get people to guess the word ‘experience’. Less effective at improving retrieval strength.

The problem with this is two-fold. For the learner, it feels easy. They get all the gratification that comes with being successful, but with none of the learning. However, in the moment, it’s so straightforward that it doesn’t share in the same desirable difficulties that are necessary for memory. It’s easy, and easy is nice.

On the other hand, it’s also problematic for the educator, because it looks like students are learning. They’re getting questions right, and you get to walk away feeling like a really amazing teacher, until they get to the assessment and all of that knowledge has disappeared into the ether. It creates the feeling of competence without actually building competence.

The issue with learning, and particularly retrieval as part of the learning process, is that for it to work, it requires some degree of challenge: a desirable difficulty. The only issue is that difficulty rarely feels desirable, and so we’re inclined to try to avoid that. Just in the same way that I might try to avoid increasing weights at gym or doing another rep of an exercise because it’s simply easier to not do that, we also take whatever opportunities we find to avoid cognitively engaging in such a way that we have to experience any kind of difficulty. Our brains don’t like thinking7.

This means that learners like to rely on surface cues or prompts that alleviate any need to think more deeply, whilst still achieving what looks like success in the moment. All the gain, with none of the pain! Unfortunately, the memories created by this are often less durable and, frankly, probably won’t stick around all that long. Even if they do, there’s a good chance those memories will be tied closely to that specific prompt that they’re rendered effectively useless. I don’t know about other teachers, but my exams don’t contain questions with hints like “starts with P, ends with hotosynthesis”.

On the other hand, when there is effort involved in the learning process, our brains better establish the conditions to store that information in our memories. Although an important caveat on desirable difficulties as part of retrieval is that, although challenging, the retrieval must still be successful, meaning it can’t be too difficult. I have taken to describing this to students using ‘The Goldilocks Zone’.

My current, not-totally-perfect-but-pretty-good-for-what-I’m-using-it-for comparison is the following:

Imagine I’m trying to build up my biceps. If I try to do so by lifting a piece of paper, which I would find easy, then I will eventually build up my biceps, but it will be remarkably inefficient.

On the other hand, I could try to build up my biceps by lifting a car, but I wouldn’t be very successful, and so again, I likely wouldn’t improve my bicep strength at all and might just fatigue myself.

The sweet spot sits somewhere in the middle: challenging enough that effort is required, but achievable enough that success is still possible.

Learning works in much the same way. Too easy and we run the risk of performance illusion with no retention, but too hard and we overwhelm learners with no success. The goal is the Goldilocks Zone: retrieval that feels effortful enough to strengthen memory, but not so difficult that it collapses into frustration. This is at least mildly inconvenient, because the Goldilocks Zone doesn’t feel anywhere near as enjoyable as the ‘too easy’ zone.

However, it is something I show to my students to remind them that we want to operate in the Goldilocks Zone, and does serve as a nice reminder that sometimes we have to put in at least a little effort to get a particular outcome.

visible learning trick

‘Pics or it didn’t happen’ logic.

This is the mini whiteboard’s number 1 enemy8. It’s the idea that learning is directly proportional to the amount of writing and/or bits of paper students take home in their bags each day.

To be fair, there’s a chance you may not have encountered this one, unless perhaps you also have students on mini whiteboards for ~90% of your lesson. Often this manifests as the student attempting to copy down every single question in their exercise book, alongside completing them on their mini whiteboard, or concerns that students are coming home with only a handful of new questions or examples written into their notes books each day. In both cases, the rationale is that learning is contingent on documentation and evidence of student effort, and is supplemented by arguments that the student might need to go back and refer to their notes to study. Let’s tackle both of those, but we’ll start with the latter.

Notes should be a succinct, correct summary of the core concepts covered. In order for notes to be something students can refer back to, they need those notes to be enough of a summary that they can find the core learning they need to review it. Likewise, those notes need to be correct. There’s little point in hinging study solely off incorrect responses. That’s of course not to say that there’s no point in using incorrect examples as part of the teaching process, but if I’m trying to revisit a worked example to review something, that’s going to be far more effective if that example is actually correct.

And that’s it. Students don’t need a 400-page book containing every single thing they’ve studied or worked with as their core notes.

Although, if you do disagree with me, I can offer you a time saving solution! If 400-pages of comprehensive notes is your style, please consider just handing your students a textbook, novel or whatever your subject’s equivalent is, because that will do largely the same thing for way less of the time and effort9. You’re welcome.

Don’t worry, I hear you.

But, Caiti, couldn’t we just have a separate notes book and exercise/response book?

Yes, and, honestly, this is my preference personally. Although my students would rarely touch their exercise books in the classroom. Instead, these would be used for homework and maybe sometimes for class tasks. This isn’t because they’re not writing stuff down. It’s because they’re writing that stuff on their mini whiteboards.

So, let’s tackle that second idea: students should, for some reason, keep evidence of all work they’ve done in the classroom.

If you think they should, I would firstly ask, why? For what reason does a student need to keep a permanent record of every single question or task they’ve attempted in class? What are they going to do with that? Some light bedtime reading of their own attempts at topic sentences or algebra, perhaps?

And even if you have got a good argument for the above, have you met students? Because I can say with confidence that I have seen students throw out worksheets before having even left the room they completed them in. That’s not even to mention the skip bins filled to the brim with student exercise books and rotten apples at the end of each school year. I’m telling you, most of them aren’t keeping all the paper you give them. There’s likely an environmental piece in here.

To be very clear, I am not anti-writing. Students should absolutely write things down in class. I find that extremely valuable, and as someone who avoids student laptop/tablet use in the classroom, except in very exceptional circumstances, my students do quite a bit of writing. However, that writing can happen without the need for them to turn it into a running historical account of their lesson successes and failures.

Mini whiteboards are useful precisely because they make thinking visible in the moment. They are a tool for being able to conduct whole-class formative assessment in the moment to inform teaching. My rationale for their use is often me considering whether it would be more beneficial for me to know instantly how all my students have gone on a question or task than to not know. Unsurprisingly, there are very rare times when it would serve my instruction more to be oblivious to my students’ understanding.

Knowing how my students go in the moment matters because that is how I teach responsively.

However, there are other reasons I would preference mini whiteboards over endless book work in many cases. On Monday, Carl Hendrick kept coming back to the question:

A mini whiteboard certainly decreases that chance if used correctly and for appropriate tasks10.

It increases participation. I know for every question who has attempted it and who hasn’t. While I do have a no-opt-out rule, sometimes students get distracted and miss a question. This is fine, because I catch that within 30 seconds, rather than with book work where I might be so busy inefficiently helping an individual with a problem that 20% of my class is also struggling with, that it takes me 3 minutes to realise. Basically, they don’t miss a lot. Full participation. Almost all time on task. And they do a whole lot more than they’d do if I set them textbook or worksheet questions for 40 minutes.

It allows me to control pacing of the lesson, teaching my students to work quickly so that they will hopefully feel far less time-pressured on exams. If they’re used to working quicker in class, then suddenly the time-pressure of exams is far less daunting.

They tell me when to move on to a new concept or increase the difficulty - desirable difficulties, if you will. They tell me how to adapt my instruct to either add prompts or start to ask deeper questions to encourage deeper thinking. They literally give me the information I need to constantly adapt my instruction to best suit the learners in front of me.

And perhaps the biggest inconvenience for critics of mini whiteboards: learning is not determined by how much physical evidence survives afterwards.

Sometimes the most important thing a student did in a lesson was think. Wild concept, I know.

busyness trap

Surely all this activity means something educational is occurring.

One of the easiest mistakes to make is confusing student movement with thinking. I spend good portion of my weekends outside walking along the bay, and I can assure you that for a lot of that time, there’s very little useful thinking going on for me.

In our classrooms, this looks like students talking, cutting things out, moving between pretty coloured stations and pastel highlighters - all telltale signs that something deeply intellectual is unfolding.

Maybe it is. I guess it depends on your learning objective. Because activity and learning are not synonyms, and a ‘busy’ (read: hectic) classroom doesn’t mean a classroom conducive to learning.

Students need to actively be thinking about something - cognitively engaging with it - in order to even have something to remember. A student can also be incredibly busy whilst thinking remarkably little.

There’s merit to a classroom that involves a relatively structured routine and that may not look ‘busy’ to an outside observer. What I mean by this is a classroom can be calm and still, and still have students actively completing relevant work. Desks can be in rows, not stations, with students cognitively engaging with the content in such a way that they’re actually learning it. Obviously, a classroom where students are playing ‘dead fish’ is likely not busy enough for whatever you’re trying to teach, but the classroom also doesn’t need to resemble a circus tent.

Learning isn’t proportional to how many different activities students do in an hour. It is, however, dependent on what students were thinking about in that hour.

preferences trick

I learn best when…

Put your hand up if you read the start of that sentence and your mind immediately went to “I learn best when I listen to music”. Because, trust me when I say, I desperately wish that were true. Unfortunately, the evidence is not particularly on our side with that one.

We have a tendency to conflate enjoyment with success. If we enjoy something more, that must mean that we will get more out of it. That’s not necessarily true - although it’s not necessarily totally untrue either.

The issue is that enjoyment doesn’t lead to learning. The two may coexist, but it’s not the causal relationship that we wish, for the most part. Enjoyment is not the mechanism by which learning occurs. Something feeling pleasant, comfortable, or personally preferred does not automatically make it more effective. In fact, we are often so drawn to the familiar or enjoyable, that we will sacrifice any kind of effortful thinking to do so11.

In fact, this particular trick tends to show up in a few familiar ways.

One of those is learning styles: a frustratingly persistent neuromyth that has been debunked so many times and yet, makes so many comebacks. Let me take this moment to remind you that

“You’re not a kinaesthetic learner; you just like HPE.” - Caiti Wade, sometimes out loud to real people.

Part of why it’s so popular is because people genuinely do have preferences that we’re drawn to. Of course we do. Some people prefer pictures, others podcasts, some silence or lo-fi beats, and some like cold beverages on a Friday afternoon. Unfortunately though, those preferences don’t necessarily translate into better learning.

In fact,

“Children are more alike than different in terms of how they think and learn.” - Daniel Willingham

So whilst we know that students have different preferences to each other, the mechanisms by which they learn are far more similar than they are different. This is an inconvenient truth for the preferences argument.

What it comes down to is the idea that preference is not evidence of effectiveness. Enjoyment may support persistence or willingness to engage, but it does not facilitate learning itself. In fact, the cognitive processes that underpin durable learning are not especially concerned with whether we liked the experience. Unfortunately, learning is a little more demanding than that.

the great learning catfish

Everyone has an opinion on school, but learning is a bit of a catfish.

Unfortunately, learning is inconvenient. And what looks like learning or feels like learning might not actually be learning.

The problem is that learning requires a level of cognitive engagement that doesn’t necessarily feel good to us, and we don’t want to do that. It’s challenging, and working through that takes at least some level of perseverance that can be a pretty tough sell to a room full of teenagers.

This is precisely why the tricks mentioned in this post feel so persuasive. Ease feels like success, visibility feels reassuring, busyness feels productive and preferences feel enjoyable. And each one of those feels far better than the thinking required for learning, which, by contrast, is effortful, inefficient and uncertain.

Terrible news, frankly, for anyone hoping effective learning would mostly feel effortless and validating.

It is easy to fall into the trap of doing what feels kind for students in the moment and giving in to preferences, stepping in with support a little too quickly or prioritising fun over rigour. However, that decision isn’t really kind in the long run.

Too easy, and we risk removing the very thinking that strengthens learning. Too hard or too little scaffolding, and students disengage, become overwhelmed, or simply stop trying. What we are really aiming for is something closer to the Goldilocks Zone: enough challenge that thinking is required, but not so challenging that success isn’t possible.

Because whilst learning will be effortful, it should not be hopeless.

The good news is we don’t have to rely on intuition alone. We know a surprising amount about what helps students learn - most of the time. And, to add to that:

There is motivation there to support our students to be successful through effective instruction. The goal is, then, not to make learning effortless, endlessly entertaining, or even immediately validating, but rather to create classrooms where students think hard, but succeed often.

The best outcome is not having students avoid effort altogether, but supporting them to know that effort, challenge and success can and should coexist.

Share

You won’t be surprised to hear that I won’t link to specific posts or comments here, as I’m capable of engaging in a discussion of ideas without feeling the need to resort to invalidating another person. Where it’s relevant, I’ll sometimes include a source or link when taking down ideas, but I take care when I do that to ensure that my criticisms are of the argument and not the person. As such, even if I reference something someone has said, their name will rarely be mentioned in my discussion - which it shouldn’t for anyone genuinely criticising an idea over an individual.

Likewise, I am also conscious that, as someone with a reasonable following, it would feel deeply irresponsible - and frankly a bit cowardly - to post in ways that risk spurring on a witch hunt toward another individual. If I’m choosing to challenge ideas, then the ideas should be robust enough to stand on their own, without requiring an audience to turn disagreement into a pile-on. Of course, that leaves me open to accusations of arguing with strawmen, but I’d much rather take that path than I would publicly diminish another person - even if it puts me in the firing line for critique myself.

So by all means, accuse me of strawmanning here. I think I’ll survive, especially knowing that I’ve chosen to take a stance that doesn’t target someone else.

Which brings me back to the argument itself. Why is low-variance such a controversial topic in the first place? Surely we want students to have equitable access to high-quality instruction?

Often, the discussions about low-variance hinge off a manufactured false dichotomy: the idea that low-variance means no variance, or that evidence-informed practice means making professional judgments redundant. If that were the case, I would not be in support of a low-variance curriculum. I have written about the need to be responsive in teaching, as well as how explicit instruction done right will probably be some of the most responsive teaching you will encounter. To be clear, this isn’t me taking a soft stance on a low-variance curriculum either. Just a friendly reminder of the following from my last post:

At face value, that seems like a difficult proposition to openly oppose. After all, who is genuinely out there arguing that students should be placed into an instructional lottery, dependent on some combination of socioeconomic position and luck?

I suspect that’s why the more public critics of low-variance instruction love a good false dichotomy. It’s far easier to shift the argument from “should students receive reliable access to effective instruction?” to “do you want teachers to become robots?” Likewise, it’s far easier to generate fear in teachers by telling them that low-variance will strip them of responsiveness, professional judgment and a level of individuality, than it is to consider how all of those things feature in, and are maintained within, a low-variance instructional model.

At the risk of making a sweeping generalisation1:

No serious advocate for evidence-informed instruction is arguing that teachers should stop adapting, responding, building relationships, using judgement, exercising personality, or adjusting to context.

So, let’s put some fears to rest. Without resorting to taking down any individuals.

myth 1: it’s robotic

The fear of loss of individuality.

One of the more common criticisms of low-variance teaching is the idea that it turns teachers into robotics - identical clones of each other, programmed to hit ‘next’ on a well-animated PowerPoint. It’s the fear that shared resources, agreed instructional approaches or evidence-informed routines somehow strip away individuality.

To feel as though your personality, humour, relationships, or professional expression are so fragile that they are stifled by a worked example or shared slide deck is a very sad perspective to have, and I feel for anyone who has experienced whatever level of micromanagement is takes to feel that way, because that’s not in the spirit of effective instruction at all.

Surely our individuality lives in our expression and delivery of the content. For me, it’s in my tone, or silly catchphrases, or the stories or humour I bring to the lesson. It’s me drawing love hearts around fractions when I teach how to add them, or referring to the integral sign and variable of integration as a “happy couple”. It’s the little class jokes like saying “morno, gents” as a greeting or, my signature move, hitting the whiteboard. That’s the individuality. It’s not whether I personally authored every single question or task my students complete.

So why, suddenly, is the idea of sharing effective instructional approaches framed as an attack on individuality? The hidden assumption here is that low-variance means standardising the whole teacher. I’ve openly said previously that I thought this would be the case when I first started teaching in a low-variance environment, but I realised very quickly that all it meant was having well-sequenced, carefully curated examples and questions prepared for me, and that the time I saved not creating these, now got to be better spent deciding how I was going to deliver these resources. My individuality wasn’t stifled because it was never tied to a resource. It had always been tied to the delivery.

Now we know my view is that showing teacher personality is far less important than fulfilling my ethical and professional obligation to teach well, but I do think for the sake of alleviating some fear, it needs to be acknowledged that individuality can and does exist in low-variance teaching. The hidden assumption in the myth that low-variance teaching turns teachers robotic is that it means standardising the whole teacher. It doesn’t. What it standardises is access to teaching sequences and activities that benefit students, whilst maintaining the individuality of the delivery.

What it potentially threatens is the occasional beloved activity that teachers designed themselves and enjoy running with their classes - regardless of the instructional quality. But maybe that’s not a bad thing. If the strongest argument for an activity is “but the students really enjoy it”, I’d wonder what the purpose actually is. It’s not to say there should be no joy in the classroom, but if it’s an activity that the team have agreed serves little instructional merit and the fun of it isn’t worth interrupting an instructional sequence for, then it’s probably not something that should be run.

As a practicing classroom teacher, there are days I have the desire to throw on Shrek and let my class relax for 70 minutes while I mentally switch off. Honestly, my classes would probably love it in the moment too. However, I know that by making that decision, I’ve disadvantaged my students compared to a teacher who chose to press on with content that day. That’s reason enough for me to stick to the term planner. But that’s more about making sure my students aren’t casualties of an instructional roulette, and if, as a team, we decided that the mathematical or otherwise benefits of watching Shrek were important, then we should all do it. Individual expression is appropriate and should even be celebrated - but only when it doesn’t limit the instructional quality that students receive. That feels less like oppression and more like basic professionalism.

But even with these agreements around resources and sequencing of instructional activities, teacher individuality can shine through. Teacher personality can and should vary, but student opportunity should not.

myth 2: it’s death by powerpoint

The fear of being dull.

I’m sure we’ve all sat through a PowerPoint so boring that mentally doing your own taxes feels like the most exhilarating option. Of course we feel sympathetic to students being put through that. However, I’d hope you have also seen a good presentation at some point. You know, the ones that feel a little like a TED Talk, where you hang on the presenter’s every word and leave feeling a little inspired to change the world or something? Not so much death by PowerPoint now.

The difference is in the delivery.

Since we’ve spoken about how expression already and how you can put an individual spin on a slide deck, let’s instead talk about this fear of being boring in the classroom. I see this one manifest in two ways: boring delivery and a lack of lesson variety. Let’s start with boring delivery, since we’ve already touched on how energy, humour and engagement can sit within delivery of almost anything.

Consistency doesn’t automatically mean monotony. This idea certainly makes sense if we imagine the evidence-informed teacher standing motionless behind a slide deck, clicking through animations and mechanically reciting bullet points into the void while thirty teenagers quietly lose the will to live. Which, to be fair, does sound dreadful. Thank goodness that caricature isn’t at all reflective of low-variance teaching.

Instead, if delivered well, our lessons should be highly interactive. And no, I don’t mean ‘interactive’ in the sense of constantly transitioning between different activities or students cutting and gluing paper in a way that’s vaguely adjacent to the curriculum. I’m talking cognitively interactive. Any worthwhile low-variance curriculum should facilitate active student participation through frequent questioning, opportunities to respond and a constant back and fourth between the teacher and student. It’s the kind of lesson where the students do so much work that they look up at the clock and realise the bell is about to go for the lunch that hadn’t even crossed their mind because they’d been kept so busy. It’s hard to argue that that’s death by PowerPoint.

Sure, it does lack the variety of constantly changing activities, but what it lacks in transition times, it makes up for in time-on-task.

To be honest, I defend PowerPoint a lot, but I’m also not married to it as being the mechanism for presenting information. Personally, I like it because the fact I can have individual slides and can run around my classroom with a whiteboard marker and clicker in hand means I can keep my lessons interactive without being chained to my laptop screen. I like that I can quickly skip through examples I’ve chosen not to do with my class, and don’t need to worry about minimising how much information is visible at once because every slide has been curated to minimise extraneous cognitive load. Use whatever tool makes you happy. I quite like a document camera too for what it’s worth.

The next way I see low-variance being referred to as “death by PPT” is in the translation that it means every lesson is an identical PowerPoint presentation for the foreseeable future and variety cannot exist. Firstly, it can. Low-variance is about minimising variation, which could mean planning for some kind of variety. But I don’t think it needs to at all. The variety lives within the fact each lesson is so interactive and responsive that no two lessons are the same.

For several years, I’ve taken two Year 122 maths (Maths Methods for my VCE/QCE readers) classes in the same year. I have taught them with the same resources and what should be the same explanation, but if you’ve spent five minutes with teenagers, you’ll know that two equivalent lessons of mine have never once been the same. Literally even with the same resources, there is variety in the way the lesson is delivered. I suspect this is in part due to the fact the interactive nature of the lessons means that it’s so easy to have the lesson go down a different path based on student response, but also based on interactions with students. Sometimes a student makes a funny comment and we write that into our scaffolds, or sometimes I’m reminded of a semi-relevant anecdote that I relay to one class in the moment when it simply did not emerge in the other. Any teachers who have taught two of the same classes have probably experienced the same and seen how the different classroom cultures can shape the lesson differently - even if they choose to use the same resources in both of their classes (something I suspect they would do).

The point is this:

Even with identical resources, low-variance teaching does not produce identical lessons. The moment students enter the picture, sameness disappears. The lessons don’t need to become monotonous, and they certainly don’t need to lack variety. If enacted as intended, any good quality low-variance resources don’t have the students sit there as passive recipients of death by PowerPoint. Instead, they are highly interactive, responsive and cognitively engaging.

myth 3: it’s pedagogical handcuffs

The fear of rigidity.

To be fair, I made up the phrase “pedagogical handcuffs”3, but I get some amusement out of the mental image it conjures. This is rather the idea that a low-variance curriculum removes the ability to respond to student needs in the moment. Of all the core arguments against low-variance instruction, this one strikes me as being a peculiar direction. Why? Just recently, the humble mini whiteboard was under attack for some reason, when its core instructional purpose is to allow for in-the-moment responsive teaching at a whole class level. And yet, we’ve now shifted to claiming that a low-variance curriculum means there’s no ability to respond to the needs of your class?

At least those of us arguing for evidence-informed practice can stick with a consistent idea of what that actually looks like.

Anyway, here goes:

You’re not handcuffed to the slides, even in a low-variance model.

When teaching using a common curriculum, there’s plenty of opportunity to respond to your class, both instructionally and relationally. I don’t know who needs to hear this, but as my students enter the classroom, I don’t refuse to greet them or ask about their weekend rugby game or congratulate them on their chess results because it’s not in the slides. If I’m circulating the room checking homework, I’m getting to know students. If a student entered my room clearly needing medical or emotional support, I’m not going to tell them to wait until the 5-minute quick ‘independent practice’ task on Slide 34 so that I can talk to them.

Just in the exact same way that a lesson plan that involves students entertaining a round table debate about causes of the fall of the Roman Empire isn’t going to take precedence over the literal health and wellbeing of the young people in that room. It’s an absurd scrutiny to suggest that the use of low-variance resources strips teachers of their ability to respond when they need to, as if a set of slides is somehow any more restrictive than a time-stamped lesson plan.

But what about instructional responsiveness? How can that be possible in a low-variance model?

I’ll say it again:

You’re not handcuffed to the slides, even in a low-variance model.

I’ve said on several occasions that, as someone who has taught in a low-variance model, I have deviated from the slide deck. Most of the time it has been me realising my students are missing some foundational skill, so blanking my slides to teach them this on the spot, but sometimes it’s been even more than this. I once saw students struggling with a topic in senior calculus and decided to completely deviate in the way I delivered this content because I knew the way I was doing it wasn’t effective.

I know, I know, that’s not very low-variance of me.

I did also immediately leave class and tell the two other teachers of the subject what I had done so that I wasn’t creating an instructional lottery between classes. And we did what any good teaching team would do. We looked at their results on the next assessment and determined my method was more effective than the one that we had already had in the low-variance resources, so we wrote it into the resources for the next year, and all retaught our classes with that same method. Had it been less effective, I wouldn’t have run it like that again. The whole point is that every decision here was done not trying to advantage one group of students over another. It was about sharing and collaboration so that all students in that cohort would have equal access to the most effective instruction - removing the instructional roulette that would exist if we taught from completely different resources.

By that measure, it’s certainly possible to deviate or exercise professional judgment where needed in a low-variance model. In fact, some level of variation is actually built into a good low-variance model. That does sound like an oxymoron.

Good low-variance resources should be created in such a way to facilitate responsive teaching. They should include checks for understanding so that the teacher has the ability to formatively assess students and either reteach or move on, based on their live formative assessment data. These checks for understanding determine lesson pace, and where more examples need to be drawn upon or misconceptions addressed to support student learning.

Because here is the reality: good low-variance teaching is not rigid. It is responsive by design. It is reducing inequitable variation while still leaving room for professional judgement, responsiveness, and improvement.

A well-designed low-variance curriculum should include constant opportunities to check understanding, surface misconceptions, slow down, speed up, reteach, extend, revisit, and adapt. The difference between this and a higher-variance model is that the responsiveness becomes collective rather than individual.

low-variance isn’t no variance

I think that’s the point that sits underneath this entire debate. There isn’t some false dichotomy here that stipulates that low-variance instruction must contain no variance whatsoever. Of course it should. The alternative is a ridiculous argument and, frankly, not even possible. The question is about whether or not there should be variation in teaching, because there should.

The real question is:

What variation do we want?

Do we want variation in humour, relationships, storytelling, personality, delivery, examples, classroom culture, responsiveness, and professional judgement? Absolutely. Of course.

Do we want variation in whether students receive clear explanations, coherent curriculum sequencing, opportunities to think, retrieve, practise, experience checking for understanding, or access teaching most likely to help them learn? No. We don’t Our students are entitled to that.

Low-variance has never been about limiting individuality or removing professional judgment. It is entirely about eliminating inequitable access to effective instruction, and by no means is low-variance no variance. It is simply the deliberate decision to reduce variation in the things that matter most for learning while protecting the very things that make teachers human.

And there we go. There’s a takedown of an argument and not an individual. That wasn’t that hard.

Share

“If explicit instruction is death by PowerPoint, then teacher autonomy is death by inequity.” - Leah Myers, Leah Myers Consulting

Every now and again I come across a sentence that cuts through educational noise in a few hard-hitting words, and this one from the incredible Leah Myers did exactly that in the last week. Don’t let the brilliantly catchy sentence distract you though - the point Leah makes is far more important than the provocation itself.

Underneath what often masquerades as internet arguments about PowerPoint or low-variance curriculum resources is a deliberate avoidance of a far more important question:

How much instructional variance are we willing to tolerate before it becomes an equity issue?

I suppose it’s far easier to fixate only on the most superficial parts of the conversation than it is to confront the inequity that emerges when instructional variance becomes too high. That does seem like a tough conversation to have.

That’s why I liked Leah’s post so much. She said the quiet part out loud. Clearly, confidently and unapologetically - exactly as we should when talking about issues impacting equitable education. Honestly, it’s probably the least we can do with those moral obligations that exist within the teaching profession.

Now if you have also had the unfortunate experience of doom-scrolling any kind of teaching social media page, you probably won’t be surprised that not everyone agreed with Leah. In fact, parts of the comment section were reminiscent of a few commentaries I had the displeasure of coming across in the last week, centred around professional autonomy. Before you come at me (don’t worry, there’s time for you to do that later), I am not against professional autonomy. I exercise professional judgment daily in my classrooms.

However, I am also strongly of the opinion that arguing about the aesthetics of instruction is a cop out, often used to avoid confronting the harsh possibility that excessive instructional variance could be creating deeply unequal experiences for students even within the same school.

students over slides

Students should not experience fundamentally different chances of success simply because they were allocated to different classrooms.

There’s a part of me that hopes someone tries to argue with that sentence, rather than just retreating into discourse about PowerPoint slides. Seriously, imagine arguing that students should have to rely on luck to have access to effective instruction. It’s a wildly regressive position to have, and yet it does seem to sit quietly beneath a surprising amount of the rhetoric surrounding professional autonomy and resistance to low-variance instruction.

At this point, it’s worth acknowledging that educational inequity exists long before students enter the gates of your school. Students’ home lives, socioeconomic status, school location and access to resources all shape students’ educational experiences and create profoundly unequal starting positions for many students.

Which is precisely why the idea of making access to effective instruction uneven as well is so difficult to justify. There are factors that are not and probably will never be in our control as classroom teachers or school leaders. That’s a sometimes-unfortunate reality that we have to face in this profession. However, we have the choice - and the professional and moral obligation - to do our best to reduce inequity in the areas we do have control over. While schools can’t eliminate every inequity that students experience outside the classroom, they can at least work to ensure that effective instruction is not another one added to the list.

so robotic teachers are the answer?

Instructional variance is not harmless when it’s the students who experience the consequences of it.

I’ve said it before and I’ll say it again: please come into my classroom and genuinely try to accuse me of being robotic. As someone who routinely receives feedback from teachers of neighbouring classrooms about my excessive enthusiasm levels, I find the caricature of evidence-informed, low-variance teaching as sterile or personality-free genuinely bizarre. We’ll come to that in detail shortly, but in the meantime, we should talk about why a low-variance curriculum is an effective mechanism for reducing instructional inequity.

The first piece to this puzzle is acknowledging the importance of effective teaching. As stated in the aptly titled ‘Ending the Lesson Lottery’ Report from the Grattan Institute:

“Effective teaching has the biggest impact on student learning outside of influences in the home.” - Hunter, Haywood & Parkinson, Grattan Institute1

There is something so deeply empowering and profoundly hopeful in that sentence. In a field where there are so many things impacting the young people that we work with that are outside of our control, we still have so much power in our classrooms. The work we put into preparing lessons, considering how we will explain an incredibly difficult concept, anticipating misconceptions or interacting with our students all matter. In fact, according to the research, they matter enormously.

And yet, despite the enormous importance of instructional quality, we routinely expect teachers to individually build large portions of this work from scratch. That is neither efficient, nor equitable. It’s at least part of why workload is a major concern, or even “unmanageable”2, for many teachers. When every teacher is left to independently construct their own interpretation of the curriculum, explanations, examples and lesson materials in isolation, instructional quality is inherently going to vary.

Picture the graduate teacher walking into their first ever solo lesson, trying to manage simultaneously connecting a projector, coordinating student behaviour and supressing their own nerves. Then let’s add to that mix a lack of experience and probable exhaustion from creating their own resources. How is their instruction going to compare to others in the profession?

Now I don’t subscribe to the view that a graduate teacher is inherently less effective than their more experienced peers - namely because I can actually name graduate teachers whose instruction is extremely effective - however, I hope you can see my point. When teachers are left to do all planning for themselves, variation will exist in the quality of the lessons they produce. Some students encounter highly coherent, carefully-sequenced instruction and others do not. Others are given lesson activities that exist largely just to ‘keep them busy’ for a period because their teacher - unsurprisingly in an overworked profession - simply ran out of time, energy or cognitive capacity to build something better that week.

Importantly, this variation isn’t a reflection of care or effort. Honestly, I don’t think it’s even always teacher expertise. Often it’s the consequence of a system that expects teachers to interpret a curriculum, create resources and plan for instruction, while simultaneously managing other elements that contribute towards unsustainable workloads.

The result of this is predictable: instructional quality will vary.

The Grattan Institute touch on this point precisely in the report referenced above:

“Many Australian teachers are being left to fend for themselves, creating lessons from scratch and scouring the internet and social media for teaching materials. This has contributed to unsustainable workloads and a high degree of variation in teaching and learning from one classroom to the next. It creates a lesson lottery for teachers and students.” - Hunter, Haywood & Parkinson, Grattan Institute

Read that last sentence again.

This is the exact rationale for a low-variance curriculum.

Not only does it reduce unsustainable teacher workload, but it also reduces the extent to which effective teaching depends on the timetable.

At its core, a low-variance approach attempts to create greater consistency in the quality of instruction students experience between classrooms. It aims to ensure that students encounter coherent curriculum sequencing, clear explanations, high-quality learning activities/resources and evidence-informed instructional practices regardless of whose class they happen to be placed in.

While lesson resources might seem to only be a small part of the teacher process, research would indicate they matter more than they may first seem:

“Research shows that when teachers use carefully sequenced, high quality curriculum materials – even if developed by others – they can boost student learning by about one-to-two months each year, possibly more.” - Hunter, Haywood & Parkinson, Grattan Institute

This doesn’t particularly surprise me. High-quality curriculum resources do far more than simply providing teachers with a few PowerPoint slides or worksheets. They embed decisions around sequencing, modelling and examples. They also reduce the enormous cognitive and logistical burden that teachers face when they have to construct every lesson, sequence and explanation from scratch.

And perhaps the time saved in not needing to build resources from scratch can instead be redirected towards other important teaching - things like considering phrasing, explanations and anticipating misconceptions. Or just sleeping or engaging in a personal life, which are also pretty important for teacher wellbeing, and likely have a reasonable impact on instructional quality.

In other words, high-quality shared curriculum resources do not remove the need for teacher expertise. If anything, they create greater capacity for teachers to exercise that expertise where it matters most.

low variance ≠ high control

Autonomy is not virtuous if the outcome is inequitable access to effective instruction.

I’ve been very much on the pro-low-variance train in this post, but I don’t think low-variance comes without concern for teachers. I’ve spoken several times3 about professional identity and the importance of any kind of change effort respecting this. At the same time, I’ve also previously posted this Substack note:

It specifically included the following:

“We can say that by neglecting effective academic instruction, we’re fostering students who are somehow holistically ‘better’, but who’s that conversation really about? The student? Or is it secretly about the teacher who doesn’t want to give up their personal instructional preferences?” - Caiti Wade

If effective instruction genuinely required me to give up some of my own instructional preferences, I would do it. My autonomy within my professional role - a role in which I am entrusted to provide students with the best possible education I can - is far less important than ensuring students consistently experience effective instruction.

However, as I often say, low-variance doesn’t immediately mean high-control. It’s not about turning teachers into clones of each other, and nor should it be. The concerns teachers have around moving towards a low-variance curriculum are grounded in fear, and some of that fear is the belief that consistency will come at the expense of professional autonomy, personality and creativity within the classroom.

I simply do not believe that has to be true. Actually, I know it doesn’t. I’m pretty open about the fact I’ve worked at a school known for having a low-variance curriculum and that I still had the flexibility to exercise professional judgment or deviate from the slides4 when it what I deemed best for my students. And if we think my personality didn’t shine through, please ask anyone who has taught in the same building as me about my lessons. Trust me, they heard them5.

If anything, working under a low-variance curriculum meant that I was learning from the incredible teachers in my team. And if ever I felt, or showed, that I had a more effective way of doing or sequencing something, we would add that into the resources for everyone to use. Built within this curriculum was significant flexibility to respond or adapt to the students in front of each teacher, and the shared resources were not enacted identically between staff. This is actually referenced explicitly6 in the Grattan Institute Report I’ve drawn upon:

“… all Maths teachers in a given year-level use the same detailed PowerPoints, but tailor their delivery for their particular classes – some classes run at a quicker pace with students completing more problems independently, while other classes move more slowly…” - Hunter, Haywood & Parkinson, Grattan Institute

Really, low-variance resources are just that. Often, they’re a set of sequenced questions, examples or tasks. Sometimes they come with particular methods to use or sample solutions. But that’s largely where it stops7. They don’t dictate the way a teacher modulates their voice or whether they choose to throw in a story about consuming too many chicken nuggets8 in their lesson.

On that note, they’re not a new idea. We often draw on the same examples, worksheets or texts as our colleagues. Maths teachers using textbook examples, I’m looking at you. The step that shifts it into being low-variance is typically just that low-variance resources are usually sequenced, rather than being standalone tasks that might be shared. If you can teach the same textbook example as thousands of other teachers and not feel as though it stifles autonomy, there’s a good chance a low-variance curriculum won’t be too much of a shift.

More importantly, a teacher can absolutely work within a coherent, shared and carefully sequenced curriculum while still exercising enormous professional judgment in how they explain concepts, respond to misconceptions, build relationships, motivate students, use humour, manage discussion and bring learning to life within their classroom.

Low-variance instruction does not need to mean identical teachers delivering identical lessons in identical ways. It means reducing harmful inconsistency in the things that matter most for student learning, while still allowing teachers to remain deeply human, relational and individual practitioners.

In many ways, I would argue that this is actually a more meaningful form of autonomy.

Rather than it being the ‘autonomy’ to give up your personal life to independently reinvent curriculum from scratch every night, maybe it’s autonomy to focus professional expertise where it matters most: the enactment of excellent instruction in front of real students.

classroom roulette

While adults argue about scripts, slides and preferences, students continue to experience vastly different levels of instructional quality between classrooms.

The reason Leah’s post, and the subsequent arguments it seemed to stir, struck such a chord with me is because a lot of the concern around low-variance curriculum resources come more from a place of fear. While those fears are valid, they can also stand in the way of something far more important: equitable access to effective instruction for students. When instructional consistency is treated as being inherently threatening, our classrooms risk becoming a game of instructional roulette, where educational opportunity depends far too heavily on luck.

There comes a point where adult discomfort has to be weighed against student outcomes.

At the same time though, I do wonder how much of this fear would exist if teachers were given the tools and training to know how they can exercise autonomy and professional judgment within a low-variance curriculum. When I read posts critiquing low-variance resources, often I wonder how much - or how little - support staff were given in implementing them.

Because it does seem that one of the more prevalent misconceptions surrounding low-variance instruction is the belief that teacher consistency and individuality are somehow mutually exclusive.

They are not.

A school can absolutely establish greater consistency around curriculum sequencing, instructional quality and evidence-informed practices without demanding that every classroom feel identical. Likewise, students can experience consistent access to effective teaching while still experiencing vastly different teachers.

Share

This week the Education Endowment Fund (EEF) and UCL Institute of Education released a major report on the effect of different student groupings in mathematics, and to be honest, the findings shouldn’t really be surprising. In summary, every way we group students simultaneously solves and creates problems for different learners.

I’m sure you won’t be surprised that I’m reviewing this article. In fact, I’ve had people message me to ask me to talk about it. The reason for the delay is really that it took me a few days to read 86 pages, summarise the key bits I wanted to talk about in any meaningful way and then construct a post - all while teaching teenage boys maths, leading pedagogy, coaching students in cross country, recording podcast episodes, and hitting up the gym. I think I possibly slept at some point there too, but I’m not even sure anymore.

One of the most common questions I’m asked when I present about mathematics teaching is whether schools should stream or attainment-group students. In fact, I was asked exactly this by a group of teachers recently, who wanted my thoughts in a single sentence. Now that is a challenge. It’s an important question and it’s the kind of question good teachers who care about their students should ask, but it’s not something I think you can adequately respond to in one sentence.

In fact, I wrote a very detailed post on that exact topic very recently right here: we need to talk about streaming.

In the above, I discuss my concerns around streaming/setting/ability-based grouping, the weird and whack research (of which there’s now more!), and present a model that seems to at least decrease some of the issues that exist with classic models. On that note, there’s probably little reason for me to talk to it again, but with new research coming out, I couldn’t help myself.

what’s new?

Before I summarise some of the key findings in the report, we should give a quick summary of the language being used:

• Mixed attainment groups: classes where students are allocated based on the timetable with no consideration given to prior attainment.

• Setting/streaming: classes where students are allocated to classes based on their prior attainment.

Likewise, I will use LPA to refer to “low-prior attaining” students and HPA to refer to “high-prior attaining” students. Although will add in the caveat that, as I said in my last post on streaming, I would typically avoid language that references “low” or “high” (and preference ‘early’ and ‘further’). Being that I am referencing the EEF and UCL paper, however, I’ll try to stick reasonably close to their language1.

Alright, let’s get to the bit you’ve been waiting for: what does the new research say?

• Overall, students in schools with mixed attainment groups typically made less progress than students in streamed groups.

• For LPA students and those from disadvantaged backgrounds2, streaming and mixed attainment groups were similarly effective.

• For HPA students, students made more progress in classes that were streamed.

But wait, that’s not new. In my previous post on streaming, I summarised the following, which basically says the same thing:

“streaming is effective for higher ‘achieving’ students, but not so much for their lower scoring peers3”

Now that alone should be very simple to follow, right?

Because based on the above, that would suggest that the optimal way to group students in mathematics would be as follows:

1. ‘HPA’ group

2. Mixed groups for everyone else

And at that point, I suppose you could just call it a day and move on, right?

WRONG.

To be fair, I once landed upon the above model having stumbled upon similar research much earlier in my career, but very quickly realised some flaws with it. We will get to those in detail, but it’s important to note that, as I put quite aptly in my post on streaming, the research is “whack”. In summary, you can find research around streaming that suggests pretty much anything you want, which is why I’m less inclined to give a one-sentence answer to this deceptively complicated question of whether to stream or not to stream.

why is the research so whack? can i trust any research??

There’s a certain level of discernment you need when reviewing any research, but particularly research involving human behaviours. This might seem like a bit of an anti-evidence stance for someone who claims to be an evidence-informed practitioner who is currently embarking on her own journey in cognitive science research, but I assure you it’s not. If anything, it’s the opposite. Taking evidence seriously means understanding its limits - and in some cases, like my own little research journey, using them to inform future research directions.

So in order to talk about this research, let’s take a quick trip back to middle school science for my non-science/non-research readers. I’ll spare you the technical language as much as possible.

When you want to test whether something causes an effect, the goal is simple: isolate the variable and make sure the test actually tests the impact you’re evaluating. Ideally, this means having two or more groups where the only thing that varies between those two groups is that one variable you’re testing.

If you want to know whether sunlight helps plants grow, you keep everything else the same: the amount of water, the soil, the temperature, the pot size and whether you let me anywhere near it4. The only thing you change is the amount of sunlight that the plants receive, so if there’s a change in plant growth, you can be pretty sure that was due to the sunlight.

That’s the ideal. Unfortunately, education research rarely gets that luxury. When we try to run experiments on student grouping, we are not just studying grouping. We are studying grouping tangled up with curriculum delivery, teacher expertise, school culture, student behaviour and task quality, just to name a few things. I once thought you could account for these factors by having one teacher who uses a script with equally distributed classes in terms of prior attainment, but any of you who are teachers will know how even the most perfectly lessons can be derailed by an unexpected student blood nose or bird flying around the classroom.

And all of that is before we even get to the problem that streaming in one school can look completely different to streaming in another.

Based on that, it’s not surprising that the research is messy. We’re testing different things in different circumstances. That’s possibly going to give us different results, and we shouldn’t be surprised by that.

I won’t deep dive into research methodologies here, because I can already feel this becoming a side-step that turns into a side-hike and eventually a full expedition into the impossible task of designing flawless experiments in education. And that’s a pit I’ve spent far too much time in personally to take you all with me5.

But, Caiti, if the research is all flawed and we can’t control variables and birds are getting blood noses while students are flying around the classroom wrecking even the most perfectly planned experiments, can we actually trust any education research?

Yes, you can. I won’t over-detail this here for fear of another potential side-step that goes too far, but there is plenty out there in the way of education research that gives us some pretty clear indications about things. However, like anything, those results can be impacted on by external factors and may have boundary conditions - which don’t necessarily diminish the results, as I discuss here and here. Over time though, we can build up a clearer picture to narrow our theories and results, which can make some of this research far more conclusive.

For instance, see below about cognitive load theory research:

“All cognitive load theory effects are based on multiple, replicated, randomised, controlled trials.” - Sweller, 2024

where does the new research sit?

The study conducted by EEF and UCL is one of the largest studies conducted on student groupings in mathematics, which, in itself, provides an interesting new piece to the much bigger, messier puzzle of streaming. However, there are some flaws with this research, even with the impressive nearly-15,000 students involved. In fact, they do openly acknowledge this within the report and the website summary, describing the ‘security’ of the new research in some cases as having a “low/moderate” rating. This diagram, taken directly from the report page, shows this uncertainty clearly:

Well. At least they acknowledge it. They do, however, make the following claim the same report:

“…we found that for high prior attaining students in schools using mixed attainment grouping made, on average, two fewer months’ progress in mathematics compared to students in schools using setting. This finding was statistically significant, meaning that it was unlikely to be due to chance.” - EEF and UCL

This raises questions about how their experiment was done. Rather than deep dive that, let’s summarise a few experimental design issues with this study.

The first is it wasn’t a randomised controlled trial where schools were randomly assigned to mixed attainment or streamed models. Instead, researchers matched schools that were already using either model, then compared them. Fundamentally, this means that there was little control over other factors, such as teaching, assessment and any other underlying differences between the schools. What is important is that the researchers did attempt to account for this:

“A pool of potential comparison schools with similar observed characteristics were then identified through statistical matching.” - EEF and UCL

That is a legitimate and often necessary methodological step. But statistical matching is not the same as experimental control. It can reduce differences we can observe; it cannot eliminate the ones we cannot. This does not invalidate this research.

The second delves into a similar line of thinking: “mixed attainment” and “streaming” look different in different classrooms, even in the same school. I work with a lot of schools and classroom-to-classroom, there’s often significant discrepancies in teaching. I could even say that there are lessons of my own that vary significantly in quality - I do my best, but sometimes I don’t quite get where I wanted to in my instruction. More to the point though, as a maths teacher who teaches both streamed and mixed attainment classes, I will make the following comment:

The relative effectiveness of my lessons for mixed attainment versus streamed classes is impacted on significantly by the content being taught.

This is anecdotal, although I’m not sure there would be much reason to disagree with this. I’ll give an example. Let’s say I’m teaching a mixed attainment group.

If the content being studied is something students have done in previous years, these lessons are typically far less effective than if the content is new for everyone. The discrepancy in retention and understanding between learners of different prior attainment levels is far wider in cases where they should have learnt that content in the past. That’s obvious - prior attainment suggests the students have previously attained that knowledge. If they have HPA, they already know most or all of it. Those with LPA may not have ever seen it before, so it’s hard to teach all effectively.

On the contrary, if the topic is something that no students have been exposed to before, my lesson will usually be far more effective for the entire class in a mixed grouping, regardless of prior attainment, because I can teach all students from the same starting point. Other factors, namely prerequisite skills, can impact student success, but overall it tends to be far better than skills they have been taught previously.

This matters because it makes broad conclusions from this study difficult to make.

Compared to some other experimental designs in education (such as cognitive load theory, as referenced above), the conclusions here are necessarily more cautious. The study can point to an association between grouping model and outcomes, but it cannot confidently tell us that one caused the other.

Other issues may have been present. The authors acknowledge that for schools using streaming, students may have been misallocated. That is, a HPA student being placed in a middle class, for instance. This wasn’t accounted for, but I would in part defend this as being part of the experiment, being that some students are likely misallocated (based on whatever criteria is being used) in most schools. I appreciate that this is reflective of real classrooms.

Again, the authors acknowledge this. They don’t over-claim here. I mention this because I am conscious of the desire to find a single answer to a question that is far deeper and would caution anyone reading this research to take it as gospel, so to speak. I, by no means, intend to invalidate the fact that this team have gone out to conduct a large study into an important question.

instructional casualties

“Dr Reid Lyon uses the term “instructional casualties” for the group of students who have difficulties learning primarily because of poor instruction.” - Del Rio, Noura, Jones & Sukkarieh, Equity Economics6

Put bluntly, instructional casualties are the students who are failed in their learning as a result of the instruction they have received. It’s an appropriately hard-hitting term, in my opinion, to name the impacts of less effective instruction.

When we consider attainment groupings in mathematics, the research by EEF and UCL tells a story that highlights the risk of instructional casualties in both mixed attainment and streamed models of grouping. That part is something I took interest in because if we want to do what’s best for our students, we need to understand the impacts of our decisions. Only then can we start to devise a possible solution.

Again, I have spoken about this previously here, so if you’re really here for an attempt at a solution, that post is probably far better than this one. Here, I want to look into some of the stories that can exist behind this new research.

casualty #1: higher prior-attaining students

The result, put roughly: HPA do better in streamed classes than mixed attainment groupings.

Why is that? Well, the report provides a possible answer:

“…we found lessons from mixed attainment schools tended to be more similar to the bottom set than top set in terms of pace…” - EEF and UCL

It’s reasonable to assume that if students are taught at a slower pace, they’ll make less progress. A streamed grouping allows students to explore content at a faster pace and often with a larger focus on extension material. In fact, this is acknowledged in the report:

“… and, although there were opportunities for challenging extension activities for high-attaining students, these opportunities frequently involved mathematics that was not directly related to the lesson content and were rarely discussed in class.” - EEF and UCL

If students are given a task and not taught how to do it, particularly if it’s not relevant to the lesson content, it’s not surprising that they won’t do as well with it as if they were taught it - or at least the fundamentals needed to access it. But in a mixed attainment grouping, what else do you do? When you cannot teach every student in that room effectively by meeting them at their point of need, you will typically try to get the LPA students over the line, so you teach to them. The HPA students have the fundamentals. They are more likely to cope on their own, so you hand them an extension task to keep them occupied, and you teach the rest of the class.

It’s not surprising that they end up being an instructional casualty of sorts here. Perhaps not in the same way that they cannot even access the core curriculum knowledge, but it would be a hard line to argue that this is optimal instruction for them.

What I find myself coming back to constantly in this is how we define HPA students, because from my work across schools, this is usually relative to the school’s own cohort. This is problematic. I detail this in my previous post about streaming with the following anecdote about two schools that have I worked with externally:

“The NAPLAN7 data for one school’s “top” maths class is approximately equivalent to the other school’s “bottom” maths class. Very similar. Statistically unremarkable. In other words, the students in those two classrooms perform at roughly the same level.

But they are not exposed to the same curriculum.

In the first school, that “top” class is taken through the curriculum at pace. They reach extension content. Challenge is built into their day-to-day lessons. Mastery of core material is an expectation and enrichment is planned for.

In the second school, the “bottom” class is labelled a “support class.” They are taught some of the core curriculum. Not all of it. Because that would be too much.

If we took these students and moved them between the two schools, they would have a very different classroom experience, and that’s the concern I have with streaming.” - Caiti Wade, we need to talk about streaming

This raises an important question about the statements made by the researchers in the EEF and UCL report:

Why do we teach mixed attainment classes at a slower pace?

Or perhaps a better question is,

Do we need to teach mixed attainment classes at a slower pace?

I’m going to say no. Hear me out for a moment here. I’m not coming at this as someone who doesn’t have experience with very mixed attainment maths classes. I’m not defending this model for grouping either. I am saying this as an advocate for responsive teaching.

When I choose to do frequent checks for understanding, I am responding to what my students can do with the content in front of them in the moment. I can say in streamed groups, I have taught ‘earlier classes’ in mathematics (classes with LPA students) who have done extension work with me in the past, in multiple lessons. This is after they have done the exact same content as every other stream in that year level, because all teachers taught using the same lesson slides. By going in with the preconceived idea that my students were LPA and therefore needed a slower lesson, I’d have never exposed them to this content. The tool that gave me the ability to do this was the mini whiteboard and some very frequent checks for understanding (CFU). This isn’t a new idea:

“[CFU] tells you when to speed up, slow down, or reteach. In reality, your students’ ability to successfully answer CFU questions determines the pace of the lesson.” - Hollingsworth & Ybarra, Explicit Direct Instruction

You were talking about streamed classes though, Caiti. How does this apply to mixed attainment groupings?

I wonder if pacing of the lesson might change for a mixed attainment grouping if the pace was determined by student success on CFU questions. Would this perhaps mirror a HPA class more than a LPA class in general? I’m not sure, but I think it’s a question worth interrogating. If my LPA classes have done extension work in the past - something usually reserved for HPA classes in a streamed model - surely that would mean a mixed attainment grouping would be capable of the same, but are they ever given the chance to do this?

As another important observation that may offer some defense to mixed attainment groupings, teacher mindset obviously plays a role in teaching. I want to explain this with what might be a surprising example.

I have also taught at some stage two maths classes in the same year level with the same course. I have taught these with the same lesson slides. One class was a HPA class and the other was a LPA class. In some instances, my LPA have outperformed my HPA class on some concepts. When that has happened, it’s always been worth interrogating, and the answer to why is often the same:

When I am conscious that a class has an earlier starting point in their relevant knowledge, I will typically scaffold more in my modelling.

This is even with me using the same lesson slides and same overarching methods - it’s just that with LPA classes, I’ll introduce more scaffolds to help meet my students where they’re at in terms of existing knowledge. As it turns out, sometimes my HPA students would have also benefitted from these same scaffolds. And those are scaffolds I likely would have given had I had more LPA students in the same class, for want of a better way of wording that. A mixed grouping probably would have been far more appropriate in these cases.

The problem you have hopefully established from the above is that the success of this model is dependent on a lot of different factors. What I do want to note very explicitly is, however, that a large spread of prior attainment within a single class does present inherent challenges. That’s a difficult thing to teach to and there’s a good likelihood that at any given moment in a mixed attainment grouping, you’re neglecting the optimal instructional for a portion of your class. How do you choose who that is?

casualty #2: lower prior-attaining students

The result, put roughly: LPA typically make similar progress in mixed and streamed groupings.

The above result is the one I think classroom teachers would find surprising. In a mixed attainment setting, how many times have you thought, I wish I could spend more time practicing this skill? Or I wish I could give more scaffolds? More guidance? How many times have you wanted to support the students struggling, but you’ve had to move on to suit the majority of the class?

It’s likely you’ve had to leave students behind in the instruction, and it’s likely there have been times you have known that a student cannot even access the basics of your lesson due to limited prerequisite knowledge, but you have little choice but to press on. Teach the majority. Try to intervene individually. It’s a challenge. So by that measure, it seems that streaming would benefit our LPA students far more than a mixed grouping. At the very least, we could scaffold more, fill in prerequisites or spend more time rehearsing specific skills. Then why does it not come across more effective in this new research?

There’s likely a variety of factors, but the one I would emphasise as being particularly harmful would be exposure to the curriculum - or lack thereof. What I mean by this is often when we group students by attainment, we then teach ‘down’ to the LPA class. That’s not done out of cruel intent, but rather misguided kindness. It’s an attempt to give the students success and not expose them to concepts that we have decided are too challenging before they even enter the room. I would summarise it as taking your students who have the earliest starting point in their knowledge and exposing them to the least content. That is a practice that runs a very dangerous risk of widening the gap between learners. The ones who most need exposure to the content get less of it, and they’re the least likely to land upon the correct answer alone.

This practice is why I say that if streaming isn’t done very carefully, I am firmly against it. To not expose students to the core curriculum is choosing students to be instructional casualties.

Interestingly, something in the EEF and UCL report may on first glance seem to sit at odds with the above:

“For the subgroup of students in receipt of FSM [free school meals], there was effectively no difference in attainment, indicating that, contrary to some qualitive evidence, students from disadvantaged backgrounds are not further disadvantaged by the practice of setting.” - EEF and UCL

For reference, the receipt of free school meals (FSM) is the criteria used to define a student as being from a disadvantaged background in the study. This is based on eligibility criteria to receive FSMs, based on factors such as family income.

Now the above would suggest that streaming is not problematic for learners from disadvantaged backgrounds. Similarly, the report outlines that LPA students also made similar progress in mixed and streamed groupings. Based on that, it would seem that my suggest that streaming can be harmful to LPA students is without much base.

However, I would hypothesise there’s more to this story. Recall the following:

“…we found lessons from mixed attainment schools tended to be more similar to the bottom set than top set in terms of pace…” - EEF and UCL

In the classes observed, the authors acknowledge that mixed attainment classes are more similarly paced to LPA classes in schools that use streaming. Based on that, it makes sense that students would make similar progress. You can call a class whatever you want; what matters is what you teach, how you teach and the pace of the lesson. If we’re saying pacing is similar, then perhaps instructionally they’re not all that different. That may suggest progress is dependent on instruction, more so than who is sat in the classroom.

Wait, did we just do all of that and land on ‘instruction is important’?? Why did I read this?

Honestly, I’m impressed you did, so thank you for taking the time. Reading the 86 pages of this report and deciding which elements to talk about took me quite a bit of time, so I appreciate you at least reading my efforts.

The quality of instruction will always be a big, if not the biggest, factor in any kind of student progress. Part of that quality is exposure to curriculum. This is something hard to consider in experimental design. Imagine me trying to determine whether LPA class teachers do actually teach less of the content. I would need several schools, and I couldn’t tell them what I was doing of course. I would also need to do observations because I can very easily tell you that what’s documented as being taught in schools often isn’t what actually gets taught. That’s probably worth pondering. Then I need some way of comparing between classes what was and wasn’t covered. It’s just an experimental design that blows out very quickly.

There’s also teaching quality too. Now before I make the next statement, I do want to say that I am a firm believer that experience in teaching does not necessarily lead to higher expertise; however, I would also never place a graduate teacher or a teacher new to my school on a LPA class if I were using streaming. Why? Because I don’t know their expertise, and the students most at risk of being instructional casualties due to poor instruction are those who struggle the most. As such, I’d always put my most expert teachers on LPA classes - although I will add in that I do try to give staff a mix of groupings overall and that I’ve been lucky to have a lot of options for expert teachers in my teams back when I was a mathematics leader.

This idea seems fairly logical. We want the teachers with the strongest pedagogical knowledge on the classes who need the most support. The HPA students can often fill in the gaps of less effective instruction, whereas the LPA students don’t always possess the background knowledge necessary to do so. Despite this, I can say from experience that some schools allocate their most ‘expert’ teachers to the HPA classes, either based on teacher preference or on the idea that the more challenging extension mathematics needs a more effective teacher to teach it. The latter isn’t necessarily untrue; however, I would state that if I’m weighing up having a less effective teacher on my LPA class or HPA class, given extension content isn’t core, I’d choose the HPA class. A sub-par attempt to teach non-core extension content is far less damaging overall than a sub-part attempt to teach core content, even without considering the prior attainment of the learners in the room.

Interestingly, there are a lot of viewpoints on this online. Some teachers are arguing that graduate teachers, for instance, should have a wide mix of classes so they get exposure to everything, otherwise they’re not learning. On that line of logic, I suppose we could also teach our 5-year-old students (prep/reception) calculus alongside how to count so they get exposure to everything.

Fine, I’m being a little facetious. But I do think there is merit in gradually building up the skills of an early career teacher or even a teacher of lower ‘expertise’8. Just as we would hopefully gradually release responsibility to our students, the same could be done for our teachers to let them master some of the basics first before moving them onto more advanced pedagogical techniques.

I’d give the example here that at some intentionally unspecified point in my career I have had to devise a plan for teaching a Year 129 maths class who were missing a significant portion of the Year 11 content due to external factors. Imagine teaching using integral calculus to find areas under graphs to students who don’t know how to substitute a variable into an equation, let alone graph a cubic function. Tricky. For me to fill in that level of background knowledge while still covering core content required some very intense planning, formative assessment and consideration about delivery of everything. And seriously, I mean everything. Even having taught this particular Year 12 maths subject many times before, with my experience in maths leadership and reasonable knowledge of education research, this was a challenging one - a challenge I would not give to a teacher without a lot of relevant expertise.

Back to teacher selection. It’s important. So much so that it’s mentioned in the EEF and UCL report:

Table 21 shows the decisions around how teachers were allocated to classes in schools using streaming. There were 68 schools included in the above. Of those, 13 say they place subject expert teachers with the LPA classes typically. 18 say they use teacher strengths to inform decisions around classes, which may be similar.

As always, the judgments that inform these could vary significantly and I wondered looking at that how much teacher preference or interest might have been conflated with teacher “strengths”. I know the above asks how many schools allocated teachers to classes based on their personal preferences, but I’m also not sure what that kind of question would produce. Namely because some teacher preference are probably taken into consideration, but some are quite likely not. If a graduate tells me they want Year 12, I would be unlikely to give it to them immediately, purely because I want them to get a bit of experience and I want to assess their ability to teach Year 12 before letting them do it.

From all that, we again land on not a whole lot of anything, but also a lot of everything. The streaming debate seems largely unsolved still; however, we can likely agree that instructional quality, exposure to content and teacher expertise are important considerations if we want to avoid our LPA being instructional casualties.

casualty #3: malcolm in the middle

The result, put roughly: Streaming was more effective overall than mixed attainment groups.

The above finding says that if we considered all student progress together, streaming is the better pick. Woohoo! Problem solved!

So why didn’t you lead with that, Caiti?

Well the EEF and UCL answer this question themselves in their own summary:

“This finding has a low to moderate security rating and should be interpreted with some caution.” - EEF and UCL

As the authors caution, it’s not worth using this as your silver bullet for answering the streaming question. I’m actually not fully convinced that even exists.

In fact, one of their other main findings appears to contradict the above finding:

“There was some evidence of a narrowing attainment gap in maths between low- and high- prior-attaining pupils in the mixed attainment schools, driven largely by lower progress among higher attainers rather than greater progress among low attainers.” - EEF and UCL

Putting these findings together:

• Streaming is more effective for student progress

• Mixed attainment is more effective for lessening the attainment gap in students

These are both important goals of mathematics education of course. I’m sure some people reading the report will take either of those as standalone statements to argue for either streaming or mixed attainment groupings, but we know considering the bigger picture is important. The authors even specify that the lessening of the attainment gap in mixed attainment groupings was driven more by HPA learners making less progress than they would in a streamed class - I’m not sure that our argument that mixed attainment is ‘better’ should hinge off the fact that some students learn less so the gap between learners decreases. To be clear, I’m not claiming the authors argue this either. I appreciate their juxtaposition of that important piece of information beside this finding. I didn’t even have to dig to find it, and I like that transparency.

Now should we stream then? Well, if you’ve listened to anything I’ve said, if you’re not going to exercise extreme caution when you do it, I wouldn’t.

what now?

Reading this piece, it has highlighted again a bigger piece in the maths education landscape. While the EEF and UCL report centred on streaming versus mixed attainment groupings overall, the more important question sits underneath it: who becomes the instructional casualty of the model we choose? And is there a way we can avoid instructional casualties altogether?

In a mixed attainment grouping, we run the risk of losing the LPA student who is perpetually lost and has to sit in every single lesson with a reminder of this, but if we stream, we run the risk that they won’t be exposed to all of the curriculum. On the other end, mixed attainment grouping can mean neglecting the opportunities for engagement with extension concepts for our HPA learners. The uncomfortable truth that this report highlights is that every grouping model can create tradeoffs.

I appreciate that it’s opened this conversation up more. In my previous post on streaming, I speak about ‘grouping for instruction’ as a model that mitigates some of the concerns discussed in this post. For those who have not read that post (pls do), it’s not just an alternative name for streaming. There’s a bit more strategy around it.

At any point in mathematics, we need to meet learners where they are on their journey and support them to build knowledge from that point to get to wherever we’re aiming for them to be. Meeting them at this point means being adaptive. It cannot simply mean equal access to the same classroom. It has to mean access to instruction that allows every student to succeed.

A grouping model alone isn’t going to do that. However, maybe there is a combination of grouping and instruction that can. At minimum, this would mean:

• all students have access to the same core content

• teacher expertise is a consideration in supporting LPA students to meet the core curriculum - remember that they have further to go in their knowledge development than their peers with HPA

• instruction needs to be responsive enough to maintain both challenge and access simultaneously, and be adaptive to the evidence in front of the teacher and not preconceived views of student ‘ability’.

In his keynote address at researchED Ballarat, David Didau distilled great teaching into two rules:

Perhaps this is where this conversation should centre. Ultimately, grouping is a structural decision, but any grouping will only be as effective as the instruction received. The grouping conversation is interesting and important, but it shouldn’t eclipse the bigger question at hand:

How do we ensure every student gets the instruction they need to best succeed in the curriculum before them?

Maybe it’s instruction, not grouping, that we should be shining the light on.

If you’ve gotten to this point and haven’t read my post we need to talk about streaming, you could always do that now!

Share

Forget the narrative that the goal of evidence-informed instruction is to improve scores or school rankings on league tables. Surely if that were the case, it would be far more advantageous to gatekeep these ideas. No, seriously, it would. If I really wanted so badly to get an academic edge on other schools, I wouldn’t openly promote the ways I’m doing that in spaces where anyone can access it - mostly for free too, being that I post the bulk of it right here. Again, that’s not unique to me. I think you’ll find those vocally advocating for evidence-informed instruction have also figured this out and would just keep the secrets of effective practice to themselves if their efforts were centred on improving their school’s results. In that respect, that narrative seems like a silly one for anyone to even attempt to argue.

Typically, criticisms of evidence-informed instruction skirt around the equity question, likely because it is difficult to argue against the proposition that all students deserve access to effective teaching. This week, though, I read a critique that took a different approach: arguing that equitable access to effective teaching smuggles in a competitive model of education. This wasn’t so much done in the sense that evidence-informed practice is all about league tables and school rankings, but rather that widespread adoption of effective instruction assumes that schooling is fundamentally about competing for future pathways, such as tertiary entrance.

I have chosen not to link to this particular Substack at this stage, being that it is being run by an individual, and I’m conscious to not spur on an online witch hunt. But I consider this argument one worth engaging with because it reflects a line of critique that positions evidence-informed practice as being results-driven, and that warrants a conversation.

the weird and whack logic of equity as competition

At the heart of this bizarre critique is the claim that hidden beneath the push for evidence-informed practice, such as explicit instruction, is a model of education founded on academic competition.

Please buckle up for a moment while I break down the flawed logic used to support this position for you.

As context, proponents for evidence-informed practice and explicit instruction, such as myself, would argue that by choosing not to use explicit instruction, some students are given an advantage over others. Basically, if we leave students to inquire or develop knowledge themselves, an inherent advantage is awarded to students who already have higher relevant background knowledge - often those from higher socioeconomic backgrounds. In fact, I make this very argument in a previous post:

“The decision to not use instructional techniques that evidence suggests are most likely to support learning is choosing to limit students’ opportunities. It can shape what knowledge they encounter, how securely they learn it and what pathways post-school remain open to them. And the consequences of that choice aren’t shared evenly.

Students with access to tutors, resources and support outside of school might find ways to compensate. Students without those supports don’t get that option. For them, the classroom is the opportunity.” - Caiti Wade, “this isn’t an academic school”

I stand by the above. Obviously. But apparently, some read an argument like this as a kind of Trojan horse: not a case for equity, but a covert attempt to transform schooling into academic combat, or education into a ruthless sorting mechanism for university pathways and social status.

I’m sure when you hear people point out that less effective teaching does not affect all students equally, your first thought is probably also: ah yes, neoliberal ranking systems and credential combat. Right?

That’s a slight exaggeration of their position, of course. To be fair, their argument isn’t quite so academic ‘Hunger Games’.

Put plainly, the weird claim is that by suggesting that limiting access to evidence-informed instruction gives some students an unfair advantage over others, we are implicitly framing education as a competitive enterprise, where the prizes are university entry and the likes. From there, the argument extends further: if schooling is treated as a multi-year training program for competitive academic outcomes, then the push for explicit instruction must also be implicitly devaluing non-university pathways. Basically, those in favour of evidence-informed practice see university as the only valuable outcome of schooling and any other pathway is a second-tier alternative.

Bet you didn’t know when you came to my post today that you were going to be accused of undermining every post-schooling pathway that doesn’t involve a university campus. Because I certainly didn’t see that one coming when I was sent the argument in question.

Now if you’re like me, you’re probably experiencing some warranted confusion at exactly what logic underpins the above argument. After all, it seems to concede two fairly important things: first, that explicit instruction does produce stronger academic outcomes than many alternative approaches; and second, that not using it can reduce equity by disproportionately advantaging students who are better equipped to compensate for weaker instruction. Surely, we want to do what’s most effective and equitable for our students, right?

Well, this author’s broader philosophical claim is this: education is primarily about individual choice, and by prioritising effective instruction, we risk positioning educational success as being solely about academic outcomes. Their argument, in part, is that while some students may want this, not every student desires a university pathway. Now that part is absolutely fair and, frankly, obvious.

But they proceed to take this argument further by comparing schooling to elite sport. In high-performance sporting environments, a narrow definition of success makes sense because these athletes have chosen to be there. School students, by contrast, do not opt into education in the same way. They suggest that by embracing evidence-informed instruction, schools are accepting a narrow definition of success founded only on academic outcomes and tertiary entry, and therefore, are risking becoming coercive.

There’s a lot there, so to try simplify this, this weird and whack logic goes something like this:

evidence-informed instruction improves academic outcomes → prioritising it makes academic success the goal of schooling → schooling is then just training for university → non-university pathways are devalued → and because students didn’t choose that goal, the whole thing risks becoming coercive.

It is at least an interesting2 argument. It is also very wrong. Figures.

how we teach is not why we educate

The first and most obvious problem with this argument is that it confuses instructional method with educational purpose. Choosing to teach explicitly is not the same as choosing what students should value, pursue, or become. It is simply a decision about how best to support them to learn. It’s not that deep - explicit instruction is a method, not a mission.

I probably don’t need to say too much more here, but I will.

So, the argument is that by teaching effectively, which will improve overall academic outcomes if done correctly, this positions academic outcomes as being the primary goal of schooling. But what if it’s not about that? What if it’s just about doing something as best we can?

For instance, I’ve taken to learning to cook well. I have zero intention of ever being a chef. It’s just a skill I want to have. It’s a flawed position to assume that by supporting students to do the best they can academically, schools are forcing them down a particular pathway. You can be good at something and choose not to pursue it. Choice does not disappear because competence increases. If anything, competence expands the choices available.

That, to me, is the strange inversion at the heart of this argument: the suggestion that helping students become more capable somehow narrows their future, when in reality it does the opposite. It opens doors. Which ones they choose to walk through is still entirely their choice.

Perhaps the author’s deeper concern is not really about explicit instruction itself, but about what schools might neglect in the pursuit of it. That if we become too focused on effective teaching, we risk narrowing the broader educational experience. Fine. That at least would be a coherent concern. However, that’s not an argument against effective instruction; it’s an argument against imbalance, which I could (and intend to) flip the switch on as well. Let’s play that game.

Firstly, there seems to be a far-too-common false dichotomy at play here: the idea that if a school prioritises effective instruction, it must necessarily neglect everything else. Forget wellbeing or sports or arts or vocational pathways. Don’t even mention pastoral care. If you choose to pursue effective instruction, you might as well not even bother with anything outside of that. Because we all know you cannot possibly do more than one thing well simultaneously, right?

Make sure you’re sitting down when I say this next part because it may come as a shock. I use explicit instruction in my maths classes. I also choose to coach a sport at the school I work at. Turns out you can do both.

I know, I know, it’s hard to hear. Imagine a school being capable of teaching well and caring deeply about wellbeing, sport, the arts, and alternative pathways. Wild.

In fact, there are countless schools that excel in numerous domains. As it turns out, schools are capable of holding multiple priorities at once. They only become competing priorities if we design them that way, so maybe choose not to? That feels like the obvious choice.

Secondly, let’s flip the switch the other way. If a mainstream school chooses to neglect its academic instruction and instead only prioritise other areas, doesn’t that also create a problem? The part that struck me most about the argument at hand was that of student choice. Not all students will choose a university pathway, but some will. And they don’t deserve to have their choice neglected either.

It’s interesting because providing effective instruction doesn’t limit student choices. A student can still finish school and choose a non-university pathway. However, not teaching effectively can close doors for students. Not because students chose that either. Because that option was taken away from them.

The argument presented was that school should be about choice, so surely our responsibility should be to maximise it? And given we can easily acknowledge that schools can teach effectively and still excel at other ‘non-academic’ (for want of a better phrase) priorities, why not do all the things?

I often say that teaching effectively is one easy thing we can control that takes pressure off in other areas. When we teach well, students don’t go home and struggle through their homework. They don’t need to rely on external tutors that they may not have the means to access or help from family. Nor do their academic results get driven purely by the hands they were dealt in life. To teach well is to maximise the effectiveness of the instructional time given. What happens outside of that isn’t always evenly distributed and often isn’t in our control. The classroom is the one place where schools can intervene in that inequality.

educational hunger games

If we suggest advantage, we are declaring competition!

No, seriously, that was an argument made. I don’t really get it either.

At this stage, I will preface this by saying that depending on your schooling system, there probably is some kind of competition built in. In Australia, students receive an ATAR3 at the end of high school, which is a ranking used for university admission. That is a competition. Although what you might note is that it’s a score and not a method of instruction.

That’s the part of the game we don’t get to decide as educators. It’s built into the system. But the argument wasn’t about that part of the system alone - it spoke to the idea that by stating that poor instruction disadvantages some students, we’re establishing education as being a competition of outcomes. This should be easy to address:

Recognising that some students are better positioned to survive weaker instruction is not creating competition. It is recognising inequality.

Whoa, that was easy.

I can go further though. Recognising inequality and choosing to take steps to reduce it through better instruction is not an endorsement of competition. It is an attempt to make the system less dependent on luck, background knowledge, and support outside the classroom.

If anything, refusing to address those inequalities does far more to preserve competition than an instructional technique ever could.

there is no neutral teaching

I want to hold the magnifying glass up to that idea of choice some more. The argument is made that students aren’t choosing to be at school, like athletes might choose to compete at a high level, and therefore shouldn’t endure an academic coaching of sorts. One of the stranger assumptions sitting beneath this argument is that by stepping back from explicit instruction, schools somehow become more neutral - less coercive and more open to student choice.

But here's some hard truth: there is no neutral teaching.

Every instructional decision reflects a choice: what knowledge matters, what gets prioritised and how effectively - or not - that content is taught. More to the point, those choices carry consequences, and it’s rarely the educator who carries the cost of that choice. It’s the students in our care, who might not even know what choice they want to make in that respect yet. And, as we already know, that cost is rarely distributed equally.

Students with stronger background knowledge, more support at home and greater access to external help can often manage weaker instruction. Students without those supports aren’t always afforded that luxury. That’s where classroom instruction fits in.

So if the concern is around coercion, maybe we should be asking the following:

What is more coercive? Teaching students well or quietly limiting what they can access later by failing to teach them well now?

ethics and equity, oh my

If I’m honest, I wanted to speak to this piece because of what sits under it. Like many posts that are fighting against evidence-informed instruction, they position the alternative as somehow being more ethical or respectful of student choice. I struggle with that position.

The students who are most likely to become instructional casualties are those with the least support outside of school. In the one place where we have the ability to make a difference in this piece about equity, there are educators who choose not to. Teaching well doesn’t force a pathway or certain results. It provides students access to the best possible education in the one way we can have control to reduce inequity.

Recall that the author of this argument acknowledged the effectiveness of explicit instruction at producing academic outcomes, so this isn’t an argument about whether it is effective or not. This is an argument about whether we are willing to deliberately withhold the very instructional approaches most likely to help students learn, all in the name of preserving a version of education that feels philosophically cleaner.

Because I’m not.

Share

No, really. It was absurd1.

The staircase above, which has ~250 stairs according to some strangers on the internet2, leads down to Frenchman’s Beach. In a lapse of judgment about my own relative fitness, I started the walk down to the beach. I got part-way down and only then decided to look at the number of remaining stairs, which resulted in a horrifying realisation: there were lots of them. At this point though, it seemed silly to go part-way and then stop. Why only do something by halves3?

So, I committed to the bit and walked down to the beach.

Why am I telling you this? Because it would have been just as easy for me to stop halfway, change the plan and go back up the stairs. I’d have still had a beautiful view of the beach, still gotten in some steps and still been able to say I’d been for a lovely morning walk; however, I wouldn’t have gotten the full benefits of making it all the way down to the beach. Some things aren’t designed to be done halfway. They only work as intended when you follow them through.

You know, like explicit instruction.

explicit instruction has a place: the classroom

I’ve addressed this recently, but there’s been a shift in the criticisms4 of explicit instruction that have moved to something similar to the following:

“Explicit instruction has a place.”

Yeah, it does. The classroom.

This stance is one where the explicit instruction critic acknowledges that explicit instruction is sometimes effective, often by stating it’s great for developing procedural knowledge in novices. This argument usually goes on to say that it doesn’t help develop conceptual understanding or problem-solving skills, before arriving at a collection of vaguely defined claims about “real world” learning or “holistic” education.

Often the argument can be summed up as the following very-broad pedagogical sequence:

1. Explicitly instruct

2. Let students inquire so they become masterful problem-solvers.

Now I’m sure some people might say that I’m oversimplifying the sequence, and I am. Perhaps, the argument is that there is more of a continuum from explicit instruction into inquiry - you know, something of a gradual release of responsibility.

Whoa. Hold the front door. Isn’t that just explicit instruction?

Because if we were to consider what myself, and my other science of learning pals5, were referring to as ‘explicit instruction’, that entire sequence is built on a combination of instructional techniques that gradually release responsibility to the learner such that they can independently problem-solve.

So to summarise, these critics are really arguing for explicit instruction, and we should all go home now. However, if you still want to go in for a pedagogical showdown, please saddle up. We ride at dawn.

they mean well; they’re just confused

The confusion seems to come from an annoyingly persistent conflation of ‘explicit instruction’ with ‘modelling’. That is, some people still think that explicit instruction = lecturing. Gee, I was once very guilty of this. Let’s flash back to my initial teaching degree days, where I made the following two (what I would now describe as incorrect) statements in assignments for my teaching degree:

• “It should be noted that explicit instruction should not be the sole teaching style. Students respond to a variety of pedagogical strategies, including inquiry learning.”

• “Several studies showed that explicit instruction is highly effective at improving student outcomes. This is important when teaching foundational content to students6.”

Yikes. Someone go back in time and scare young me by showing her these posts pls.

I was commended for that section in my paper by the lecturer marking it, rather than called out for my obvious conflation of modelling and explicit instruction. Perhaps this lecturer didn’t know the difference between the two themselves and so couldn’t correct my misconception.

It’s not surprising then that this view still permeates all of the social channels we use to have our pedagogical battles, so to speak.

To make sure we’re on the same page, let’s do some brief definitions here then.

Explicit instruction:

“…the defining feature of explicit instruction is that, for novice learners, concepts are fully explained, and procedures are fully modelled before learners are asked to apply those concepts or procedures. Significantly, this working definition does not preclude the possibility of learners completing open-ended problem-solving tasks.” - Sweller, Zhang, Ashman, Cobern, & Kirschner (2024)

Typically, guidance around effective explicit instruction would go further to stipulate that tasks should be broken down into small steps and students should be given several opportunities for practice, including independent practice. Take the following from Rosenshine’s Principles of Instruction (2012), for instance:

• “Present new material in small steps with student practice after each step.”

• “Ask a large number of questions and check the responses of all students.”

• “Require and monitor independent practice.”

In fact, that active involvement of students practicing after each step, embedded formative assessment and responsive teaching are all what form an explicit instruction lesson, and draw it distinctly away from purely lecturing at a group of students. Previously, in my post on explicit instruction, I included the following image that highlights that the majority of lesson time is spent asking students questions and responding to that data:

Further underpinning explicit instruction is the gradual release of responsibility7. Gradual release of responsibility can be defined as follows:

“an approach to teaching that emphasizes a gradual shift in responsibility from the teacher to the student. For example, whereas a teacher may take most or all of the responsibility in guiding students through a complex and/or new task…, the goal of [gradual release of responsibility] is to guide students toward an increasing level of independence in completing the task” - Salehomoum, Revelle, Duke, & Pearson (2022)

Essentially, it’s the gradual transition from fully guided practice to independent practice. Just to make mention of the role of gradual release of responsibility within an explicit instruction framework, please see below from the NSW Department of Education, for instance:

“Gradual release of responsibility is an explicit teaching strategy. It is not linear as teachers move between modelling, guided and independent practice throughout lessons. Teachers make decisions based on student understanding and readiness. Students may move 'forwards' and 'backwards' in this process.” - NSW Department of Education

Importantly here, the reference to the fact this movement isn’t necessarily linear draws upon the formative assessment and responsive teaching elements of an explicit teaching instructional sequence, mentioned above.

To summarise, this means that explicit instruction is not a component of a lesson. It is the lesson. Or rather, I might describe it as being the entire sequence of instruction that’s used within a lesson, underpinned by the science of learning, whereby the teacher explicitly models all relevant concepts to students. This means that explicit instruction isn’t something that I can just do for 15 or 20 minutes of my lesson time. It is my entire lesson.

Modelling:

Modelling, or explicit modelling, is part of the explicit instruction sequence. Modelling is the teacher-led demonstration of a skill or concept before students need to attempt this themselves. This involves breaking tasks down into manageable steps, using worked examples, providing examples and non-examples, and ‘think-alouds’ during demonstrations, to name a few techniques. This publication explains this idea well.

This involves explicitly demonstrating or telling students what you want them to know. In the words of Zach Groshell:

“Just Tell Them” - from the title of his book ‘Just Tell Them: The Power of Explanations and Explicit Teaching’

Modelling is an important part of the explicit instruction sequence. In fact, it’s the key defining factor; however, it’s not the only piece to explicit instruction. Going back to my own image above, it’s just one very small portion - one that must be present, but it’s not the whole show itself.

the difference

Modelling:
“Here’s how to do it.”

Explicit instruction:
“Here’s how to do it.”, followed by a series of steps to gradually release responsibility based on formative assessment of student mastery.

¿porque no los dos?

Why not both? Why not have explicit instruction and inquiry in the one lesson?

Explicit instruction is the whole instructional sequence. Inquiry is the whole instructional sequence. And you can’t have two wholes8.

This is where it comes back to definitions. With the definition above of explicit instruction, the goal is to move students gradually through stages of learning to get them to a point where they can complete independent practice and move into non-routine problem-solving. Some might describe this end ‘portion’ of an explicit instruction sequence as being inquiry, and that’s fine. In fact, if we’re referring to inquiry as being purely something students can pick up and do, I’m totally on board with that. The idea of transitioning students into being able to generalise and adapt knowledge is literally part of The Instructional Hierarchy:

In that case, it’s really just semantics, right? If we’re all arguing that lessons should involve explicit modelling of all core concepts, opportunities for practice, faded guidance, regular whole-class checks for understanding, responsive teaching and a gradual release of responsibility with the aim being to move students towards independent work, then hoorah! We’re all fighting for the same thing and our pedagogical standoff is over. Honestly, call that technique whatever you want, because I don’t even care about the title if you’re just going to teach effectively.

However, this doesn’t seem to be the point being made in many of these mysterious unnamed internet posts I see that are trying to discredit explicit instruction. These posts typically focus on the idea of having ‘explicit modelling’ elements within inquiry lessons, removing the features that make an explicit instruction sequence so effective. Nathaniel Swain makes reference to this point exactly:

“if we are embedding the explicit part somewhere in the mix of an inquiry or problem solving mode of teaching (no matter how clearly we try to do this), I would argue this does not constitute the “explicit instruction” that aligns well with the science of how students learn.” - Inquiry learning or explicit teaching - balance, Dr Nathaniel Swain

I agree. An inquiry lesson cannot contain ‘explicit instruction’ portions. It may contain explicit modelling, or it may contain multiple opportunities for practice; however, these two teaching methodologies are inherently different. The phrases ‘explicit instruction’ or ‘inquiry learning’ are used to describe two different pedagogical sequences, not just individual strategies that can be picked up and used as one may please - although both contain several strategies. With these definitions, it means that elements of each cannot sit within the other.

fence-sitting

Don’t sit on the fence too long or you’ll get splinters. I didn’t make that quote up, but I did think I did when I was on a stair-less walk thinking about this post. Turns out a lot of people have had bad experiences having been sat on a fence, which honestly just strengthens my point that you probably shouldn’t do it.

It sounds nice to sit on the fence and argue that a lesson needs variety. To be honest, I agree that lessons need variety, but I suspect what we define as variety might differ slightly. For instance, lesson variety could mean a little bit of voice modulation throughout, or covering new concepts each day, or even bringing in a totally different creative, but probably ineffective, learning experience each day so that students never have to learn about fractions by seeing an actual fraction9. Likewise, we could definitely describe making rumballs in class to teach Business Management as lesson variety:

“What student wouldn’t think about smashing biscuits or mixing in condenses milk and say to themselves ‘oh yes, this is a process to transform our inputs to our final output, just like what Mr Brien taught us last week’? Turns out all of them wouldn’t think that.” - Shaun Brien, Ghosts of Lessons Past Pt 3

However, as Shaun very clearly stipulates above, not all variety is effective variety, no matter how delicious.

Variety sounds like one of those things you need. It’s an appealing concept, so it’s hard to argue against, which is what makes fence sitting so easy. When you consider how responsive teaching works though, it’s hard to argue that there’s not variety in that lesson structure. You’re literally adapting your lesson in the moment to what you’re seeing. It’s so much variety that even the teacher doesn’t know exactly how their lesson will go.

Personally, I also find students get plenty of variety when you just teach them new content well every day. This is something easy to do when you teach them well in the first place. You know, learning can be engaging10, and even more so when that learning has been effective. Our job isn’t to come up with a new way of presenting learning to students each day or changing our lesson structure each period for the sake of it. It’s to teach well.

Part of the fence-sitting side here that argues for bits of both explicit instruction and inquiry learning tries to draw upon supposed limitations of explicit instruction. It relies on claims that students need to be able to produce knowledge independently, problem-solve and develop the ability to adapt their knowledge for the real world - and of course we can’t argue with that. It’s right. However, those aren’t arguments against explicit instruction. Those are exactly what explicit instruction, and specifically gradual release of responsibility, works towards. The whole purpose of explicit instruction is to provide the most efficient and effective guidance to move students from novice learners into experts who can adapt their knowledge across contexts. You don’t get that outcome by doing a bit of everything or changing techniques part-way through. You get that by sticking with the full sequence.

a pedagogical showdown?

I spoke at the beginning about making my way down an absurd number of stairs at Straddie over the weekend, and how I would not have experienced the full benefits of doing so had I stopped part-way and turned back.

This is not dissimilar to what happens when only portions of an explicit instruction sequence are taken and used - often just modelling alone. Without the opportunities for practice, formative assessment, responsive teaching, gradual release of responsibility and everything else that makes up the explicit instruction sequence, individual strategies will only get you so far. It’s like stopping part-way down the stairs and never actually making it to the beach. The power of the explicit instruction sequence is the culmination of everything altogether.

So, is this really a pedagogical showdown? A battle of explicit instruction versus inquiry learning?

Because if we’re all aiming for students who can apply knowledge independently, solve problems and adapt their knowledge, then we’re not arguing for different outcomes. We are, however, arguing about how to get there.

And on that point, I’d argue that inquiry learning and explicit instruction are fundamentally at odds. Not because one values independent practice and the other doesn’t, but because they represent entirely different instructional approaches. Neither approach is just a collection of strategies or techniques that can be slapped together however one pleases. On both ‘sides’, there exists a best practice that outlines how individual decisions and strategies are sequenced, and why this is done. This means that under these definitions they can’t really be blended and they can’t coexist in the same lesson.

To say that ‘best practice inquiry’ contains some explicit instruction just cannot be the case, when explicit instruction is the entire sequence.

In some ways, yes, this is a pedagogical showdown. Just it’s not a battle of preferred outcomes; it’s a clash of approaches. Specifically, it’s a difference in how we think learning happens and how we choose to get students there, and those cannot coexist in the same lesson.

And in case you were wondering which side I would take, I’d follow the staircase all the way down to explicit instruction. Every time.

Share

It’s kinda like when influencers team up to ‘collab’1 (as they call it in the industry) on social media, but it’s actually just two mini whiteboard enthusiasts with Substacks on school holidays, and is objectively far less cool. Lots of humour, some non-examples and a Simple Model for Mini Whiteboard use (which is actually really cool: I love it a bit).

It’s the final week of Term 1 in Queensland this week and while I love my job to bits, the break will be a welcome reprieve. There’s something refreshing about an early morning walk along the beach, knowing that I’m working off a schedule of my own choosing, with built-in time to reset from the tiredness that seems to accumulate across a busy - albeit enjoyable - term.

I was recently chatting to a friend, who asked a question I get a lot: how I manage challenging behaviours with students when they do arise. I said I make a concerted effort to always start a conversation with a student from a place of kindness or understanding. That doesn’t mean lowered standards by any means, but it gives the conversation the best chance of going somewhere productive, regardless of what consequences will happen. While it’s certainly not always necessary, sometimes I find myself internally thanking myself for starting the conversation there, particularly in those moments when a student shares context that makes you quietly grateful you led with kindness. We get that a lot in life. You can come down too hard, only to realise that the person you’re interacting with was dealing with more than you initially knew.

It’s a classic case of the fundamental attribution error: we have a tendency to over-attribute someone else’s behaviour to personal characteristics, rather than considering the context around it - which is often the opposite of how we explain our own behaviours. When we go in too hard on a particular behaviour or act of what looks like defiance, we’re implicitly treating that action as being a character ‘issue’. By leading with kindness, this holds space for situational explanations before jumping to dispositional ones.

While it sounds like a nice strategy, I will admit that as the end-of-term tired kicks in, it takes far more conscious effort to maintain. However, I actively try to do it because, more often than not, it produces some kind of positive outcome, even if the behaviour was merely the result of a ‘poor choice’. That’s probably not surprising though. People are consistently more likely to engage positively in corrective conversations when met with empathy and understanding than to judgment and confrontation.1

However, like anything that involves human behaviour, it won’t land perfectly every time. There are far too many variables for that - timing, context, prior experiences or wording, to name a few. Generalisations around anything involving people share that same situational nuance, but it doesn’t mean they don’t work and it certainly doesn’t mean they’re not worth trying.

it always works, except when it doesn’t

Last week, I posted most likely. most students. most of the time. In this post, I stipulated that evidence-informed practice is grounded in the following idea:

That’s a point I standby strongly and is the reason I choose to use evidence to inform my practice. With that said, I would also state that evidence-informed practice works for all students.

Whoa, Caiti, that sounds like A LOT of contradiction.

Sure, it might sound like a contradiction, but it’s actually a distinction between principle and practice. I am aware that at this stage, it’s just as easy for someone to stop reading this and take my two quotes above, strip away all context, place them side-by-side and claim they’ve won the internet or something. In fact, to save them the time, I’ll do it for them:

Evidence-informed practice is about choosing what’s most likely to be effective for most students most of the time. Evidence-informed practice also works for all students.
- Caiti Wade, conveniently decontextualised.

Alright, now that we’ve gotten that out of the way, let’s talk to this point. Initially, I’d planned to include this distinction in my previous post, but I already surpassed the email length with Simpsons quotes, so it gets its own post. If you’ve not already read that post, I’d strongly recommend you do before getting into this one, because I do build on points made in it (remember, this was originally one post that I’ve chosen to split).

This isn’t a contradiction. It’s a distinction between two different levels of claim: principle and practice2. These require significant discussion, but to give a brief overview:

At the level of principle, evidence-informed practice works for all students. That’s because it is grounded in how humans learn, and that’s not something that magically changes because you’ve entered a different postcode. At the same time though, at the level of practice, no single strategy will work perfectly for every student, in every moment and in every context. Even with sound underlying principles, what happens in the moment is impacted by a whole range of variables, such as prior knowledge, attention, timing, delivery and how windy it is outside (the latter of which remains a largely unaccounted-for variable in education research). This does not mean that evidence-informed practice doesn’t work.

I want to take this brief opportunity to remind of us where else this same logic applies. Gravity. It works for everyone. Sometimes its effects are impacted upon by other forces, but gravity still holds, even if its impacts aren’t always consistent. Of course, you could always hit back with, “but what if you’re floating in a vacuum in the middle of space…”

Well, you would be dead, but I admire the commitment to trying to edge-case your way out of fact.

Likewise, there are medications that work - except when they don’t. There are factors that can impact upon the effectiveness in the moment or for that individual. Take drug interactions, for instance. However, in other fields, we wouldn’t consider that argument sufficient to state that a particular drug doesn’t “work”. In fact, medical professionals would hopefully continue to look into drug interactions, boundary cases and factors impacting upon effectiveness to maximise the chances of successful administration, rather than hitting back with, “well, drats, back to the drawing board”. You know, that thing people like to do in education.

We have a well-established understanding of how the human brain learns biologically secondary knowledge. While that research will continue to be refined as we learn about further ‘interactions’ that exist (such as prior knowledge, for instance), it provides a clear evidential basis for some instructional techniques, such as explicit instruction.

Realistically, we could argue that anything works until it doesn’t. The point is this: the existence of variability doesn’t invalidate the principle. It simply introduces additional factors to consider and inform judgments in practice.

level 1: the principle

Explicit instruction works for all students.

Stating that explicit instruction works for all students isn’t a claim of perfection, but a claim of universality.

This claim is grounded in what we know about human cognition and learning. Humans, regardless of background or context, are subject to universal underlying architectures of memory and attention. In my last post, I used the following quote from Daniel Willingham:

“Children are more alike than different in terms of how they think and learn.” - Daniel Willingham, Why Students Don’t Like School

That’s worth expanding on. If you’ve not read it, this 2024 paper by Sweller provides a far more detailed discussion of some of what I will reference here; however the key point, grounded in cognitive load theory, that I will reference could be summarised as follows:

The major modifiable cognitive difference between learners is differential knowledge held in long-term memory, not differences in the processing architecture itself.

That is, the way new biologically secondary information is acquired or processed is the same for all learners. Perceived differences in learning or knowledge acquisition are the result of what knowledge learners already have in their long-term memory. Typically, this is referred to as “expertise” within cognitive load theory. In fact, in the article linked above, Sweller states:

“That memory store is the source of expertise and as discussed in more detail below, possibly the only source of malleable and alterable cognitive differences between individuals.” - Sweller, 2024

In basic language, the process by which learners learn is the same. The ‘differences’ we perceive in learning aren’t due to the cognitive process, but rather the existing knowledge the learner has.

Basically, this framework asserts a universality of architecture, not performance. Cognitive load theory, whilst grounded in the idea that the way biologically secondary knowledge is acquired is the same for all learners, accommodates the variation that exists as a result of prior knowledge. In fact, both the expertise reversal effect and element interactivity explicitly reference this idea and its implications on instruction.

Of course, you could try to claim that cognitive load theory isn’t evidence-informed, but doing so would require a re-definition of ‘evidence-informed’:

“All cognitive load theory effects are based on multiple, replicated, randomised, controlled trials.” - Sweller, 2024

Based on the above, the implication for instruction is relatively straightforward. If all learners acquire new, biologically secondary knowledge through the same underlying cognitive processes, then instructional approaches that align with those processes (for instance, ‘evidence-informed’ instruction) are not selectively effective. They are universally applicable.

This is why evidence-informed practice works for all learners. This is why instructional techniques, like explicit instruction, work for all learners. Stating that they work isn’t a claim promising perfect outcomes. It’s a claim that the technique sits in alignment with the process by which every human learns. And like it or not, you don’t get to opt out of cognitive architecture.

level 2: the practice

Choose the technique that is most likely to work for most students most of time.

Whilst the principle itself might be universal, other factors impact upon its effectiveness practice. Kinda. I’m always reluctant to say ‘classrooms are not controlled environments’, being that some anti-evidence folk like to suggest that educational research was conducted in labs and not real classrooms3, but they’re not. At the level of practice, what translates into ‘learning’ (a change in long-term memory) for each student is shaped by a wide range of variables: prior knowledge, attention, the explanations used or whether it was last period on a Friday.

So while the underlying mechanisms of processing new knowledge are consistent and consistently support particular instructional techniques, their enactment will never be perfectly uniform. In fact, the point made in the previous section around the impact of prior knowledge alone means that knowledge acquisition will vary somewhat between each learner, because their prior knowledge is always going to be somewhat different.

However, this does not diminish the effectiveness of evidence-informed practice, nor does it mean that evidence-informed practice doesn’t ‘work’ for all learners. This means that, just like my example above about drug interactions, there are factors that will both positively and negatively impact to what extent a technique is effective. Variability in the outcome is not a failure of the principle alone, but it is a reality of implementation.

This is where my ‘most likely. most students. most of the time.’ idea enters the chat. At this point, you could try to say that if the effectiveness of evidence-informed practice is influenced by so many other factors, then it clearly doesn’t work. Nice try, but that logic doesn’t hold particularly well either, and might even be the undoing of ‘opposing’ techniques (I’m looking at you, discovery learning).

That conclusion confuses variability with failure. In reality, the strength of evidence-informed approaches, such as explicit instruction, is that they are robust to the variability of real classrooms. They don’t require perfect conditions to be effective. They work despite them.

Take prior knowledge, as mentioned above, for instance. Differences in what learners already know will always exist and that will impact on new knowledge acquisition. Rather than this be a reason to dismiss an evidence-informed approach, this is the exact reason using evidence-informed practice matters. If knowing something will better support our students to access new knowledge, and they currently don’t know it, then we should teach it to them. What’s that called again? Oh yeah, explicit instruction.

Compare that to approaches where students are expected to discover knowledge for themselves. Rather than addressing differences in prior knowledge, these approaches often compound them.

On the other hand, evidence-informed practice is grounded in techniques that do work. Not because they’re perfect in the face of variability, but because they remain effective in spite of it - the very reason that they are most likely to be effective for most students most of the time.

don’t weaken my evidence stance

I am 100% for science of learning (or Science of Learning if you’re into capitalisation, but by this point, you should know how I feel about capital letters), evidence-informed practice and explicit instruction. Feel free to quote me on that too - with or without context.

This is not a matter of preference. It is grounded in principle. These approaches align with the universality of how students learn and are robust enough to withstand the variability of real classrooms. That is exactly why they are most likely to be effective for most students, most of the time.

The reason I speak to the ‘most’ rule, as I’m going to refer to it, isn’t as some fence-sitting take on evidence-informed practice, nor an admission that the evidence doesn’t hold. The ‘most’ rule is grounded in classroom reality, as someone who is still a full-time practicing teacher. I know that in any given lesson, there are a number of variables at play, only some of which I have any control over. I also know that sometimes, despite my best intentions, I don’t enact an instructional technique with the level of fidelity I would have hoped for - especially when end-of-term tired creeps in.

The ‘most’ rule reflects that reality. It acknowledges that while the underlying principles of how humans learn are universal, the specific enactment of a technique within a specific domain requires ongoing adjustment in response to what we see in front of us. This is where professional judgment comes in, as I mentioned in my last post. Although I want to be very clear that professional judgment is not a substitute for evidence, nor an excuse to put personal preference above your students’ learning. Professional judgment is about considering the variables present in front of you and responding to reality, informed by the evidence, and all of its nuance and boundary conditions and the likes. The goal should really be to understand why certain techniques are more effective than others to support moment-to-moment judgments in the classroom.

I make professional judgments daily in the classroom, but my decisions about how to respond to these aren’t based on preference or vibes. They are grounded in an understanding of how students learn and how I can teach in a way that maximises the likelihood that, despite any variability or weather events in the room that day, learning still happens. Decisions based on what’s most likely to be effective for most students most of the time, using techniques that we know work for all learners.

This framework asserts universality of architecture, not uniformity of performance. Both of these things can be true at once. Evidence-informed practice works for all students, but it will also only ever work for most students, most of the time.

Evidence-informed practice works for all students because it aligns with how all students learn, but any single enactment of it will only work for most students, most of the time.

Share

If your standard for instructional strategies is that they must work for every student, in every classroom, every time, you have ruled out everything that actually works.

With evidence-informed practice gaining traction in Australia, backed by some jurisdictions (such as Victoria and NSW), some oddly anti–evidence-informed practice organisations and individuals have been forced to shift tactics in their attempts to undermine decades of research. Instead of arguing that it “doesn’t work”, the attempted takedown has moved to “it doesn’t work for all students” or “it doesn’t work in all contexts”.

I’ll admit, I almost respect the attempt there to move the goalposts once it became clear that “the evidence doesn’t work” wasn’t going to hold up as an argument anymore. Almost.

Rather than pull from the screenshots of countless LinkedIn, X or other social media teaching posts sitting in an infuriating folder on my phone right now1, I’ll draw on this quote from Dr Rachael Jefferson, as quoted by EducationHQ, as an example:

“[science of learning] posits science really as an absolute, and it also suggests explicitly that nobody can question its authority because it is the ‘science’ of learning.” - Dr Jefferson, EducationHQ

“An absolute”, you say? That’s an interesting take on what is, quite literally, just a body of research. You know, something that only exists because new claims are tested and old ones refined.

That in itself makes it clear that evidence-informed practice or science of learning or whatever you want to call it, wasn’t ever claiming certainty. In fact, when I’ve worked with schools using the science of learning as the theoretical underpinning for their teaching and learning frameworks, that conversation has centred on something much simpler than it’s often made out to be. The term “science of learning” (or “evidence-informed practice”) is just a label for what the current body of research suggests about how humans learn and, by extension, how instructional approaches can support that.

If you put it like that, it’s a weird thing for someone to argue against. Imagine being actively in opposition with the use of research in the classroom:

“I’ll try anything - except the thing most likely to work” - what blanket opposition to evidence-informed practice sounds like.

Or, alternatively, we could draw on Adam Savage from Mythbusters to provide a quote for us:

Perhaps the confusion comes from people thinking “evidence-informed practice” is a prescribed list of strategies, rather than a body of research that helps us make better decisions about teaching. Or perhaps it comes from a misunderstanding of what that research is actually claiming.

Right now, many of these conversations seem to rest on the idea that if a strategy doesn’t “work” for every student, in every classroom, every time, then it isn’t worth using. However, the claim made by evidence-informed practice is far more modest than that.

Not perfect, nor universal, nor guaranteed. This isn’t a case of choosing between ‘perfect’ and ‘not perfect’. This is about choosing between what is more likely to work and what is less likely to work.

it’s not perfect; it’s probable.

The entire premise of using evidence to inform practice is about making probabilistic decisions based on what will probably be effective. It literally works on the idea of synthesising the current evidence base and using that to inform instruction, and yet evidence-informed practice is continually judged against a standard that nothing in teaching (or several other fields) could ever meet.

Guess what? No approach works for every student in every situation every single time. If that’s the measure being used to reject an approach, why not just reject teaching altogether? After all, as Homer Simpsons says, “trying is the first step toward failure”.

Before we decide that using an approach that will probably work is somehow a bad idea, it’s worth remembering that we accept this logic elsewhere happily. For instance, laws, and the associated consequences of breaking them, don’t work for everyone. However, they do work for most, and I’d much prefer those odds than having none at all. I’ll take that option in teaching too.

It reminds me of the aphorism, don’t let perfect be the enemy of good, which suggests that the pursuit of perfection can often hinder progress. That is exactly what is happening here. The expectation of perfection leads to the rejection of potentially effective practice, not because it’s ineffective, but because it - like literally every other technique - fails to meet an impossible standard.

The game here isn’t one of false promises that can be used an excuse to abandon an idea before even giving it a chance. It’s a game of chance, where the odds suggest that under certain circumstances, particular instructional techniques are most likely to be effective for most students most of the time.

the evidence doesn’t apply to our context

Ah, yes, the old “We’ve tried nothing and we’re all out of ideas”.

I empathise with this point though. Srsly. I have worked across a variety of school contexts as a teacher - low socioeconomic, private schools, co-educational, girls only and, currently, boys only. I have also worked with several schools externally in this time, and that has given me a very nice snapshot of just how different schools can be. The benefit of that is I’ve seen how strategies apply in different contexts, but that’s not necessarily a privilege everyone is afforded.

Previously, I have made the point that a lot of the video guidance or examples of evidence-informed practice in the classroom highlight what I’m going to call the best of best practice. In some cases, the classroom set ups are so ‘perfect’ looking that I’m half convinced we’re hiring adult actors to play students. I wouldn’t be surprised to see a video close-up of Chris Hemsworth2 engaging in a perfectly synchronised choral response, delivered with effortless confidence.

However (don’t pretend you didn’t know this was coming), if we are talking about the evidence about how students learn, that doesn’t differ between postcodes. Our brains are remarkably similar in how they work. As Daniel Willingham says:

“Children are more alike than different in terms of how they think and learn.” - Daniel Willingham, Why Students Don’t Like School

This means that dismissing research grounded in cognitive science purely because the context is different isn’t quite the powerful argument it’s made out to be.

With that said, I will jump in with a very important concession here. While the cognitive science doesn’t differ, the exact way that implementation of evidence-informed instructional techniques looks will vary a bit. In fact, in my opinion, this is the power of knowing and understanding the evidence-base. It gives you the ability to successfully adapt an instructional technique to suit a particular context - and if you know the research really well, you can also avoid lethal mutations.

I want to be clear, however. This isn’t about taking an idea and going rogue or reshaping it beyond recognition. Adaptation requires preserving the underlying mechanism that made that technique effective in the first place. If that goes missing, it’s no longer evidence-informed practice.

That might seem like an obvious point to make, but seemingly not, as indicated here:

“When theoretical ideas don’t work in a classroom, it’s not because they don’t work in practice as well as had been imagined, it’s because it’s not the right sort of implementation. It’s blame-shifting from the science, which is perfect and precise and cannot be questioned, to the teachers, who are none of the above.” - Phil Beadle on his Substack

If you’ve been around for a while, you’ll likely know that it would be a rare day for me to enter a classroom without my mini whiteboards because I know just how effective they are when used correctly. But I could absolutely remove any relative benefit of using them with a few simple moves - make them optional, not respond to the checks for understanding my students have done on them, or just let students doodle willy-nilly on them as they please3 and call it a ‘brain break’. Hey, if I felt like it, I could have my students use their mini whiteboards as frisbees. If I did those things instead of harnessing their benefits for instant whole-class checking for understanding, they probably wouldn’t be effective in any context4.

Likewise, retrieval practice is effective when students are retrieving the content. That means they should have learnt it, which is more than just having been taught it. That’s a far easier mistake to make than having students use mini whiteboards as frisbees, but also diminishes the comparative effectiveness of an evidence-informed technique.

Applying a technique to a new context is entirely dependent on understanding what makes that technique effective in the first place, and often times, the shift is cultural. It’s about teaching students how to use a mini whiteboard or building a culture of engagement. The videos of ‘best practice’ I mentioned earlier are great examples and allow you to see a technique done well, because that’s important. It doesn’t mean that’s where it will begin though, nor does it mean that’s where it will end.

Likewise, the power of having a strong understand of the research base underpinning a particular strategy gives guidance about when you might choose to use it. This might come as a big shock given my apparent mini whiteboard obsession, but I also use other methods to check for understanding, coupled with what I now know to be ‘opportunities to respond’ (thanks to Bron Ryrie-Jones for improving my practice tenfold after a single presentation - again). There are times when the mini whiteboard isn’t my go-to, because I’ve made the decision that it’s not going to have the benefit I want in the moment, or another technique might be more effective. If you know why something works, you can choose when to use it.

All the evidence base does is give us some boundary conditions around what is most likely to be effective in what circumstances, and the instructional techniques based on these are purely a synthesis of how to put research into practice.

but the “evidence” is always changing

Correct. We tend to update what we do when we learn more. That’s generally considered a strength. We also stopped using leeches to treat every illness under the sun at some point too. Progress can be inconvenient like that, I guess.

It is frustrating to need to update practice or change strategies. You can feel as though you’ve mastered something when a new nuance enters the scene, and suddenly you’re needing to refine practice. What’s the alternative though? Continuing to do something that is now known to be less effective? That’s the equivalent of tipping the same team every week because they used to be the best. They’re probably not bad, but they’re also probably not the best choice anymore.

Again, the whole point of evidence-informed practice is that the evidence is based on what research currently says is most effective. I would hope that the research continues to evolve so that I can be the best educator possible.

The current body of research is purely a snapshot of what we have currently found. Likewise, the purpose of research in general is to continue to fill gaps, expand knowledge or find boundary conditions or moderators for what the research is currently suggesting. From that, we would expect change to happen and that you’ll have to make some updates to your practice to reflect the research.

Ultimately though, the evidence is very unlikely to cause a complete 180 turnaround either though. It’s far more likely to build upon or improve existing practices. In the time I’ve been involved in the science of learning space, I’ve made changes to my practice to reflect new advancements in research, but it’s largely very incremental. And wildly? I enjoy it, because I like to constantly get better.

The evidence changes because we learn. And personally, I’m not convinced that “we’ve always done it this way” is the stronger position there.

What does the evidence reflect then if it’s always changing? It reflects what the research currently suggests is most likely to be effective for most students most of the time.

but the “evidence” hasn’t changed since 19XX

Wait, which one is it? It’s always changing or never has? It would be much easier to follow the argument if we picked one lane and stuck to it.

If evidence-informed practice isn’t being criticised for always changing, it’s been criticised for being a rebrand of ‘old-school’ teaching under a new name. I’ve seen claims it’s just a rebrand to make money, and I’d like to ask how? I feel like I should be cashing in on this rebrand as someone who regularly releases free posts on how to implement evidence-informed techniques in the real-life classroom.

To be fair, some elements will look similar to ‘old-school’ teaching. Evidence-informed practice would typically have student sat at desks in rows facing a board at the front, which isn’t a new idea by any means. And frankly, that shouldn’t be surprising to us5. I could logically assume that students facing every different direction in the classroom probably wasn’t going to support listening to a teacher at the front without having even opened a journal article ever.

A lot of the techniques that are effective will have similarities with instructional techniques that have been used historically. That checks out, because we’ve been able to transfer knowledge up to this point, so it’s not like we were doing everything ‘wrong’. With that said, while on the surface instructional techniques might look familiar, the difference is in the precision with which they’re implemented.

This isn’t purely a loose collection of strategies. Evidence-informed practice is a set of approaches grounded in an understanding of how we currently understand learning to work. The strategies we have used historically aren’t ‘wrong’ (probably), but they might be incomplete or might have use cases that deem them more or less effective than other techniques. For instance, explicit instruction is often misunderstood to be lecturing to a group of students, rather than the interactive back-and-fourth we would define it as now.

Are there elements of lecturing at students that are effective? Sure. Likewise, there are elements that would deem it ineffective. Did somebody say transient information effect?6

However, if we take that idea of explicit modelling and start to pair that with research around engagement, worked examples, formative assessment and responsive teaching, we’re hitting on some powerful pedagogical practices. The whole point of the research is not that a strategy exists in isolation, but instead creating a ‘best of both worlds’ situation. Evidence-informed practice is about harnessing the benefits of a variety of instructional techniques, using evidence to support our classroom judgments, so that we can choose the technique that is most likely to be effective for most students most of the time.

most likely. most students. most of the time. now choose a new argument.

If your standard for instructional strategies is that they must work for every student, in every classroom, every time, you have ruled out everything that actually works.

Dismissing evidence-informed instruction on the basis of you anecdotally having a story of a student where this wouldn’t be effective is a great way to give up on every other student in your care, while simultaneously admitting that you’re okay doing something that’s more likely to be less effective.

At no point has evidence-informed instruction claimed certainty, nor promised perfection. It doesn’t override momentary judgments based on experience or knowledge of the other factors in your classroom in the moment. It provides a tool-kit to choose a response based on best bets for that context or that moment in your lesson.

The alternative to that is using strategies that we know to be less effective for that given context in that given moment. The evidence gives the reasons why, and teacher expertise and knowledge of evidence chooses the instructional technique.

In fact, the decision to use evidence-informed practice is grounded in this idea:

We should choose approaches that are most likely to be effective for most students most of the time.

This isn’t because they’re flawless; it’s because they give us better odds - odds that you can improve when you use the evidence to support your own judgments and experience as a teacher. In a profession where no strategy is going to work 100% of the time, that’s not a weakness. That’s the entire point.

You can reject that if you like, but you don’t get to pretend the alternative is stronger.

Share

There were so many incredible presenters over the course of the day. I was lucky enough to see Bron Ryrie-Jones, Dr Nathaniel Swain, David Didau, Dr Ben Jensen and Dr Jenny Donovan - I know, please hold your jealousy. And that’s not to mention all of the incredible names I didn’t get to see, whose sessions I’m trying to piece together through notes that other attendees have kindly shared with me.

If you’ve been following along for the last couple of weeks, you will know that I presented a session ‘The First Five Steps: Beginning a schoolwide pedagogical shift’ - a concept that, in hindsight, is very hard to squeeze into 45 minutes. The result was me speaking at roughly double speed while covering so much ground on stage that I’m fairly confident I got in enough steps for the entire audience.

A few people who couldn’t attend asked me to share my slides. I’m always slightly reluctant to do this. Not because I’m trying to gatekeep them, but because, as someone who hates a ‘death by PowerPoint’, my slides are mostly a series of AI-generated images and visuals that supplement my explanations, which leaves them very open to be misinterpreted when viewed as a standalone resource. Anyway, I have shared them here if you’re wanting to try piece together whatever story I was trying to tell and/or attempt to understand why leadership teams around the country are suddenly setting fire to their pedagogical frameworks. The following slide is an example of one from my presentation that makes a bold claim that I worry will be misconstrued without my accompanying verbal commentary1:

Immediately after putting that slide up, I said something along the lines of: “People - teachers - teachers are people too..” (Thanks to Shaun Brien for capturing that one in his notes from the session2), followed by something to the effect of “Teachers hate change, right? Wrong.”

And that slide being out in the wild without my verbal correction feels like a really great way to send the wrong message here, so let’s talk about it. Because if you think your teachers resist change simply because it’s change, you’re probably solving the wrong problem.

700 teachers on a saturday

Before I get into the real organisational change side of things, it would be completely remiss of me not to acknowledge that 700 educators gave up their Saturday with the goal of improving teaching and learning. Those teachers clearly don’t hate change. They want to improve - something that is a real theme of this piece.

The thing is, teachers inherently want to get better. You don’t have to look very far to see that. Spend a Saturday at researchED and it becomes very difficult to maintain that idea that teachers are somehow a profession resistant to improvement. In fact, look beyond a single great conference and into your own schools. Teachers read and experiment and refine their practice. They share resources or ideas with colleagues, and try strategy after strategy to help that one student when the explanation isn’t landing. They give up their lunchbreaks when they’re needed elsewhere, because they care about that job and all of those people their work impacts. These aren’t people who hate change. These are people who want to be the best they can be for the young people in their classrooms, and these are people who change their practice or routines constantly to make sure they always do this.

So why does it feel like when a new initiative comes in that teachers hate change? Where does that feeling come from?

Heifetz and Linsky say people don’t resist change. People resist loss. For schools introducing widescale pedagogical change, the ‘resistance’ experienced is rarely about the idea itself. The concern is likely about what might be lost in the process. Setting aside issues like how the change is enacted, four broad fears tend to sit underneath that hesitation. This list is by no means exhaustive, but they’re considerations for those of you leading change.

1) loss of autonomy

One of the main concerns for teachers when large-scale change is introduced is a loss of professional autonomy. Teaching is a craft. Over years in the classroom, teachers develop routines, explanations, habits and approaches that work. This was likely after entering the classroom underprepared from initial teacher education, then working independently to develop a practice that is safe and works, based off a culmination of experience and professional judgment. When a new framework is introduced, it can easily feel as though those years of learning and professional judgment are being replaced with a script.

For many teachers, the concern isn’t the strategy itself. It’s the idea that their experience, expertise, and ability to make decisions in the moment might be reduced in the process. In my opinion, something that does not necessarily help this is that a lot of the really amazing videos that demonstrate evidence-informed instructional techniques are a little too perfect. They typically show a perfectly run class with largely compliant students, and a teacher who is measured and precise in their delivery. All great things. But the reality is that the measured, precise delivery isn’t always realistic.

For instance, I had a teacher observe one of my maths lessons this week, and afterwards they asked if I was always that energetic and happy in class (usually, yes) and if I intentionally chose to modulate my voice throughout the lesson. Over the course of 70-minutes, I moved between loud and energetic, casual and bubbly, positive praise, and quiet, slow and measured, with moments of deliberate seriousness. At one point, I did reference raising of the larynx to help students to imitate me in a choral response activity.

That’s probably not something that features in many videos of ‘best practice’, despite me being a reasonably vocal3 person in the field advocating for evidence-informed instruction. I know the cognitive science behind it. I know that if I say too many words, I can introduce extraneous cognitive load. I know that when I get distracted and tell my students a funny story, it’s not supporting the learning goal. But like many teachers, I know when to pull back.

As a teacher, I’m constantly reading the room. I modulate my voice regularly to get and keep attention, signal importance or lift energy when it’s needed. I slow things down and reduce my volume at times to support maximum focus. We know when humour or stories are needed to build support, and we know when to lock in because a concept is really hard. Most of us do this somewhat instinctively.

That’s the point though. Professional autonomy in teaching isn’t about doing whatever you like. It’s about having the judgement to respond to the complexity of a classroom in real time. It’s about the confidence that you can use your professional judgment to adjust instruction, tone, pace and explanation to suit the room and students in front of you.

When a movement to a new pedagogical framework is introduced, it’s not surprising that it can bring up fears about what this might mean for teacher autonomy. If the change feels like professional judgment is going to be replaced with rigid scripts, compliance checks or tightly prescribed routines, the fear of losing professional agency becomes very real and very valid. Why? Because teachers want to keep being the best they can be, and they know that depends on them being able to respond in the moment - something that is often tied to the idea of professional autonomy. When it feels like something that important might be taken away, hesitation is a completely predictable response.

For leaders introducing pedagogical change, this matters. If it’s communicated - explicitly or implicitly - that autonomy will be removed, resistance isn’t surprising. The leadership work here is building up the idea that a shared approach to teaching doesn’t replace professional judgment, but rather, supports it. I’ve written in detail about this here, discussing how a low-variance curriculum enhances teacher autonomy, although I’d caution you against just throwing your staff my blog post and calling that ‘change management’4. Clear instructional principles should reduce unnecessary variation in the fundamentals, while leaving space for teachers to exercise expertise that makes responsive teaching possible. Part of this is having your staff work to build up your framework in their classrooms (you know, since they’re the ones enacting it) so that it allows for the movement needed for staff to exercise autonomy still.

In other words, the goal shouldn’t be to remove autonomy. It should be to use it more deliberately.

Similarly, this is something that needs to be shown gradually. It’s going to take time to shift, but understand where the concerns are coming from so that you can work to disarm them.

2) workload concerns

If you’re leading widescale schoolwide change and your plan involves teachers simply finding more time somewhere, I regret to inform you that teachers are not currently hiding spare hours in their desk drawers.

It’s not surprising that workload is one of the most immediate and practical fears that accompanies change in schools. Even when the idea itself is strong, it’s completely reasonable for teachers to calculate what this change will cost them in time and energy. If this change means rewriting resources, learning new systems, or preparing for lessons differently, it’s a hard sell. The problem is, your change probably does involve all of the above and more.

This one is an easy one to talk to because most of us in education have been there, when we’ve been up at all hours or hiding from cooking dinner because we really want to make that great resource to support our students. It’s hard to get your work-life balance right while protecting the limited time and capacity that teachers already have to do their jobs well.

This is why when leaders introduce large-scale change, they need to carefully consider what the work will look like for teachers. Good change removes work as often as it adds it. Unfortunately, for a good pedagogical change to occur, there is usually an initial wave of unavoidable increased workload. In that case, leaders need to consider when staff can do this. Realistically.

Personally, if you have the capacity, I really like the strategy of pulling some expert curriculum writers off some classes to have them prepare common materials (when you get to doing that - it might not be your first step!). This means you have the people interested making the materials and minimise diffusion risk, while also having a clear strategy for managing workload. Previously, I have also used ‘staff days’ (also ‘student free’ days or whatever you call your school’s days when there are no students, but all the staff) to have my teams work on resources when I’ve wanted to introduce a new initiative. I literally set a list of tasks for staff to do, like writing new questions on prescribed topics, and give them time to do it. It’s not rocket science. When I’ve wanted my team to do something well, I figure I should put in place the supports to ensure that happens.

I suppose you could ignore that, because I recognise it can be expensive to pull staff off classes. However, you need to keep in mind that if change only ever feels like more, it’s very hard for teachers to believe that it will eventually lead to better.

3) fear of incompetence

This one is likely to be less visible, but arguably one of the more powerful fears when change is introduced. Teaching is public work. Every lesson is performed in front of a room full of students and when techniques don’t land, there’s nowhere to hide.

When a new pedagogical approach is introduced, there’s often an unspoken concern: what if I can’t do this well?

In part, this is shaped by the way many teachers enter the profession. Initial teacher education regularly leaves teachers feeling underprepared for the classroom, so practice is built through experience. Over time, trial and error, reflection, and support-seeking gradually turn into something that feels reliable and effective. And then along comes a new technique.

Hopefully you can see why that might raise concerns.

For teachers who have developed ‘pedagogical survival strategies’ that exist purely to help them get through a lesson from start to finish, the last thing they need is to introduce something that feels like it might destabilise what is already working5. For more experienced teachers with strong classroom practice, a new technique can feel like taking a step backwards. Even if the long-term outcome is improved practice, the short-term reality is often less fluency, less confidence, and more uncertainty - likely for both the teacher and students. That’s not an easy position to step into for either of these teachers, and certainly not in front of students.

So again, this isn’t teachers resisting change or hating change. This is a very human response to the risk of temporarily feeling (or being) less competent.

For leaders, this matters. If change requires teachers to feel less capable before they feel more capable, then support, modelling and time to practice aren’t optional. They’re essential. I can’t emphasise enough the importance of trialling and modelling approaches in real classrooms, ideally at your school. It means preparing students for new techniques at a whole-school level to support your teachers in implementing these, and creating opportunities for low-stakes practice. And it means creating opportunities for teachers to experiment and build routines that work in your context before rolling out the strategy at a broader scale.

Subscribe now

4) threat to identity

The final main fear sits underneath everything else: a threat to professional identity. A change in teaching practice doesn’t just ask teachers to do something differently. It can ask them to see themselves differently.

Over time, teachers build a sense of who they are in the classroom. This is built on the kind of teacher they are, how they explain things, the ways they interact with students and the techniques they valued. This forms an identity for teachers built on years of experience, reflection and small moments of success that reinforce this is what works for me.

To go back to a quote from why are you so defensive about getting better?:

If I’ve built my identity around having exciting activities for students or creative ways of delivering content and I’m being asked to be more explicit in my instruction, that doesn’t just feel like a shift in practice. That feels like a shift in who I am as a teacher. The resistance that surfaces here isn’t about rejecting improvement so much as it is about holding onto a version of yourself that has, up until now, made sense.

And when you might have to face the possibility that what you’ve been doing isn’t as effective as you believed, that’s a confronting realisation. Because for most teachers, part of their identity is not just that they are teachers, but that they are good teachers. They care about their students, they work hard and they’ve invested time, energy and thought into their practice. They do their best and make sacrifices to be the best possible teacher they can be. So when the message - implicit or explicit - is that what their doing isn’t good enough, it’s not surprising that people try to protect their identity.

In order to build the conditions for teachers to engage with change, leaders need to recognise the identity and talent that staff bring to their classrooms. This means acknowledging the effort, care and intent behind current practice. It’s about avoiding deficit framing and being deliberate in ensuring that change is communicated as being focused on refinement and not on filling gaps in practice.

The work isn’t to convince teachers that they’re not doing well enough. It’s to build up conditions that support refining practice, without asking teachers to let go of who they are in the process.

stop fighting resistance

Teachers don’t hate change.

What leaders often interpret as resistance is usually something else. It’s a signal about what teachers are worried about. Good leadership doesn’t fight against that; it works with it.

I’m the first to admit that there’s a temptation to respond to any kind of resistance with something to the effect of “the research says…”, but that’s fighting emotion with facts, and that’s not getting to the heart of the concerns at all. It’s a surface-level response that won’t move practice forward. Instead, good leadership seeks to understand where hesitation comes from and responds to that deliberately and empathetically. It recognises that resistance isn’t something to defeat, but something to learn from.

When leaders take the time to understand what sits underneath pushback, they are far better positioned to introduce change in a way that their teams can actually engage with it. Once you understand what’s underneath the resistance, you’re no longer trying to push change through people. Instead, you’re leading it with them.

Share

I felt so lucky to get to hear from the incredible David Didau, Bron Ryrie-Jones, Dr Ben Jensen, Dr Nathaniel Swain and Dr Jenny Donovan. Events like this are always energising, and I’ve left again with pages of notes on new things to try or papers to read.

My session focused on some of the research into implementation science and change management at a schoolwide level and how leaders can turn that into actionable steps to initiate a shift in their school - essentially turning the research into a structured process that can be adapted to facilitate change efforts. Thank you to those of you who came along!

So if you’re heading back to your school on Monday with an urge to burn your pedagogical playbook, here are the slides:

The First Five Steps

3.42MB ∙ PDF file

Download

Download

** for any images that make no sense without context, feel free to make up your own story.

I’ve seen a lot of schools recently posting launch events for their new pedagogical frameworks (or whatever slick title they’ve decided on). There’s a small part of me that’s jealous and would love to spend my time launching things, even if only for the opportunity to attend an event with a multi-course dinner in a nice dress with a glass of bubbles.

The rest of me has some reservations about this. None of them related to dinner.

Developing a new instructional practice takes time. A new instructional move requires rehearsal and refinement over time to become habitual. That’s why the idea of launching an entire pedagogical framework all at once makes me uneasy. Pedagogy isn’t a single behaviour. It’s a system of behaviours - modelling concepts, effective questioning, checking for understanding, responsive teaching. Each carries nuance, and each takes time to develop with fidelity.

If we wouldn’t expect students to master a complex concept after a single explanation accompanied with a well-branded booklet, it’s difficult to imagine teachers mastering an entire system of instructional practices after one very nice launch event, a multi-course dinner, and a glass or more of bubbles. To be honest, my concern doesn’t even sit entirely with the schools posting these events. In many cases, the launch may simply mark the culmination of months - or years - of focused work on a small number of instructional practices.

My concern is more so with how this is perceived by other school leaders looking to shift pedagogy. When launches like this become visible across the sector, others understandably take notice. And sometimes what gets copied is not the years of deliberate refinement that preceded the launch, but the launch itself.

So if you’d like to ensure your schoolwide pedagogical shift fails and poison the well for future change efforts, here are a few reliable strategies you can use.

1) celebrate before you even begin

Your first move in shifting an entire schoolwide pedagogy should be manifesting the success you want in the form of canapés and a very expensive launch event. Ideally, staff should hear about the instructional practices they’re now expected to implement sometime after the oysters and before the third glass of champagne. Because nothing prepares people for a major shift in professional practice quite like alcohol, seafood, and a glossy booklet.

Launch events are excellent for signalling to the broader community that change is happening. They are far less effective at actually making it happen.

Imagine hearing, for the first time, that your professional practice is expected to change while sitting at a dinner table surrounded by colleagues. There’s no rehearsal, no gradual build-up, no opportunity to explore the thinking behind it. Just a public unveiling of what teaching is now supposed to look like. It shouldn’t be a surprise that this isn’t the most respectful way to announce an entire shift in practice.

But, Caiti, surely the title of the event would signal to people that the plan is to unveil a new way of doing their literal entire jobs?

Perhaps, but forewarning isn’t the same thing as building the conditions for people to engage with it positively, let alone successfully.

In research around organisational change, there is a consistent theme that the early framing of a reform or change has significant impact on how people respond to it1. In this case, when the change is introduced as a public unveiling (probably for marketing purposes) rather than as a process of professional learning and refinement, it’s not likely to go down well. It shifts the conversation from asking how existing practice can be elevated into people asking why their existing practice is being replaced.

As I discussed in my last post, why are you so defensive about getting better?, teaching practice is often tied closely to identity. When a pedagogical change is framed as a replacement for existing practice, it sounds like a correction. And when you’re publicising an entire correctional playbook at an event, resistance isn’t an irrational response; it’s a predictable one.

Research on organisational change consistently shows that the early framing of a reform matters enormously for how people respond to it (Weick, 1995; Fixsen et al., 2005). When change is introduced as a public unveiling rather than a process of professional learning, it can quickly shift the conversation from curiosity to judgement. Instead of asking how a new practice might work, people begin asking why their existing practice is being replaced.

In professions like teaching, where practice is closely tied to identity, that distinction matters. When improvement is framed as correction, resistance is not an irrational response; it is a predictable one.

On that note though, if you do really want to make sure your new framework is nothing more than some nice branding and a three-course-dinner, please make sure the focus is on the event itself and not the people actually having to execute your new framework. You know what, maybe consider saving some money from the budget and not even inviting those implementing the framework to the event - now that’s an idea.

2) introduce everything all at once

Go print your full instructional playbook, pedagogical framework, or whatever impressive title you’ve given it, and get that out to staff immediately. Make sure it carefully details every single part of their professional practice and exactly how lessons will now look across the entire school. The goal here is to give staff the maximum amount of new information at once with the minimum possible opportunity to rehearse any individual practice to the level of fidelity required to actually make it work. Your staff will be grateful that you’ve given them something to work on in all that free time they definitely have.

As teachers, we would never give our students an entire set of notes for a full topic at once and expect them to read these and immediately be able to execute every skill with high fidelity. So why do we do that with teachers?

When it comes to implementing new instructional practices, there’s often the desire to change the entire structure of a lesson all at once. We look to the end goal and quickly move to embed every single instructional practice required to reach that end goal - and when you’ve been teaching the same way for years, that’s going to be a big shift for both staff and students.

Implementation research suggests that complex professional practices rarely become embedded this way. Practices develop through repeated use, feedback, and refinement over time2. In other words, people don’t hear about a practice once and immediately execute it with fidelity. If they’re not able to do this with a single instructional move, it’s far less likely they’ll succeed at completely changing the way they’ve operated in their careers up to this point.

Anyone who has spent time in a classroom knows that teaching quickly becomes a collection of habits - some good, some less so. These are the things you say and do that you don’t really have to think about too deeply. They’re your defaults or fall backs, or those little annoying phrases you say far too many times in your lessons.

In case you’ve forgotten or it’s your first time here3, I use mini whiteboards a lot in all of my classes. This means that over a hundred times a day most days, I find myself saying, “3… 2… 1… show!” in full teacher voice. You know the voice. Don’t pretend you don’t.

This is a routine I’ve repeated so many times that it’s completely automatic. I realised just how automatic it was a couple of years ago at a large New Year’s Eve event. The MC of the event handed me the microphone and asked if I wanted to lead the countdown to New Year, and as an extrovert who isn’t opposed to being the centre of attention at times, you can bet I said yes.

“10… 9… 8…” - I was smashing it!

“7… 6… 5… 4…” - at this point, I was thinking I should MC weddings as a side gig.

“3… 2… 1…” - oh no, teacher voice.

“SHOW!”

Happy New Year.

You can bet that when I can’t sleep at night, my brain loves to replay the moment I accidentally asked a couple of hundred strangers to present their answers on imaginary mini whiteboards. Really great from me4.

The point, unfortunately for my dignity, is that habits become our default. I wasn’t in a classroom setting and nobody was doing maths, but I fell back on what was automatic to me. When teachers are tired, overworked or in a high-stress situation, they’re going to fall back on their habits. Ideally those habits will be great instructional techniques, like having a consistent routine for mini whiteboard usage. If we do genuinely want teachers to form habits around high quality instructional techniques, we need to give them time to rehearse, refine and repeat those strategies until they become the default.

Habit formation research suggests to us that even simple behaviours like drinking a glass of water after breakfast can take around 2 months on average to become automatic5. I would suggest that instructional practices are considerably more complex than water consumption, and so expecting teachers to adopt an entire instructional framework at once is unlikely to be successful even if just from a habit formation perspective.

In addition though, trying to embed an entire suite of practices simultaneously means none of them get the attention required to be implemented with fidelity and to become habits that endure the ups and downs of a busy school term (I’m looking at you, Term 2). It’s great to partially embed a technique, but time needs to be given to ensure that technique is implemented with fidelity6 and doesn’t become a lethal mutation.

This means that if your goal is genuinely to shift classroom practice, the smarter strategy is the slower one: introduce fewer techniques, practise them deliberately, and allow them to become habit before layering on the next.

But of course, if the goal is to kill the reform quickly, introducing everything at once is an excellent strategy.

3) remove the teachers from the design process

The next step is to design the entire pedagogical framework centrally and present it to staff as a finished product. After all, the people who will be responsible for enacting the practices in classrooms every day don’t really need to be involved in shaping them. In fact, they’ll be incredibly grateful that you’ve taken the job of designing how they spend literally every working day out of their hands. And if that seems like too much effort? Why not simply copy and paste an existing one from the internet, rebrand it, quietly ignore copyright law, and hand that to your staff instead? Trust me: if they don’t love the booklet itself, they’ll definitely appreciate the extra attention you put into debossing the lettering on the cover.

Hold up, Caiti, are you saying we shouldn’t plan our entire schoolwide pedagogical framework ahead of time as a leadership team?

Yes, I am. This might be my most controversial take, being that this seems to be the practice I see the most.

In many schools, well-intentioned leaders decide that it’s time to refresh or redesign the school’s pedagogy. A leadership team retreats somewhere with a wall that’s actually a whiteboard, a few too many coffees, and a strategic plan, and emerges several weeks later with a beautifully structured framework explaining how teaching should now look across the entire school. Now even if these leaders were to gradually introduce the team to each instructional technique that underpins their pedagogical framework, they’ve already set themselves up to fail.

“But that wouldn’t work in our context”

If I had a dollar for every time I’ve heard the phrase above in response to a suggested instructional technique, I could probably use gold foil lettering on the cover of those pedagogical playbooks. No disrespect to the comment. It’s fair. Really.

When we show teachers research around instructional techniques or videos of teachers engaging in these, these are polished products. Rarely do videos demonstrating instructional techniques showcase the times when they weren’t executed perfectly. They don’t show the teaching of routines to students or the classroom management and expectations that go into building up strong classroom routines. They definitely don’t show when you hit Week 7 of Term 1 and have to reteach the routines to your students because they’re tired and you’re tired and everyone is dragging themselves to the finish line of term.

No, these videos show an optimal execution of a technique. And in pretty much every classroom context, that’s likely going to take a lot of time to develop, and might not even look the same as it does in the video. This doesn’t mean it’s not worth trying, but it does mean it should be adapted and recontextualised to suit the existing structures at your school, and with your students.

And that’s exactly the part that becomes difficult when the framework has already been fully designed - especially if it was designed by those who teach no or few classes.

Handing teachers a finished product and expecting them to implement it is unlikely to be successful, particularly when there’s a good chance the research or resources you’re drawing upon to support implementation didn’t come from your school. Implementation research suggests that sustainable change is far less likely when handing people a completed model and asking them to execute it. Rather, successful implementation depends heavily on the ability of practitioners to adapt new practices within their context while maintaining the core intent of the strategy7. In other words, the goal isn’t perfect replication. The goal is contextualised execution.

And the way you meet that goal? Involve some teachers in the design process.

In fact, the most successful shifts in schoolwide practice typically involve processes where the frameworks were designed from the ground up8, under the guidance of a clear vision from leadership. This means that change efforts are more likely to succeed when people are involved in the design process, as compared to being handed something from the top-down. This practice increases teacher buy in and ownership, but also allows for the framework to be experimented with and recontextualised within your own school environment.

The risk here is of course around diffusion. What if staff move the framework in the wrong direction? Well, that’s where leadership comes in. Leaders set the direction, introduce the ideas and support experimentation. Teachers are responsible for refining what these ideas look like in classrooms.

This could mean pitching an instructional technique, building a shared understanding and language, and then trialling this to develop pictures of how this looks within classrooms at your school. It’s built by the people who need to use it.

One technique at a time. Seriously. I know it’s tempting to go about this in some kind of divide-and-conquer fashion, and start allocating different staff an instructional technique each to refine and build, but that will send you straight back to Step 2 in this post. Not a good move, because now you’re going backwards in a post literally designed to talk about all the things you shouldn’t be doing.

We also do genuinely want our staff to collaborate on developing these techniques. We want to consider different subject areas and make sure there are collaborative teams of staff to support each other to develop structures that support context-specific fidelity. And then, we want to document that somehow.

Over time, those shared refinements and accompanying documentation begin to form something coherent. In fact, if you iterate over this enough times, you might even end up with something that resembles a pedagogical framework.

The difference is that by the time it has a fancy name and some gold foil text on the cover, it already works, it’s already embedded and it’s already been adapted to your context.

However, if you prefer a more authoritarian approach, by all means design the entire thing without them. After all, the people expected to enact the practice every day are the least qualified to help shape it.

Subscribe now

4) pitch your pedagogy as correction

The only reason you need to roll out this new snazzy pedagogical framework is because your teachers weren’t doing a good enough job in the first place, so really they could do with the help. Make sure that message is crystal clear when you introduce it. After all, nothing motivates professionals quite like discovering that the way they’ve been doing their job for years was fundamentally flawed and that the solution has finally arrived in the form of a laminated diagram and a well-branded booklet.

The fastest way to trigger resistance in any profession is to threaten competence. Teaching is deeply tied to professional identity, so when improvement is framed as correcting what teachers have been doing rather than refining it, the response is rarely curiosity. It’s protection.

And that reaction makes sense. Most teachers have built their practice over years, often with limited guidance, inconsistent preparation, and a lot of trial and error. In that time they’ve figured things out. They’ve developed routines that work, explanations that land, and strategies that help their students learn. When a reform is introduced as though none of that exists, it sends a very clear message: what you’ve been doing doesn’t count.

Once that message lands, the dynamic shifts immediately. Instead of asking, “How could this strengthen what I already do?” people begin asking, “Why is leadership assuming I’ve been doing it wrong?” The conversation moves away from improvement and toward defending professional competence. At that point, you’re not getting anywhere rolling out that new framework.

Ironically, the opposite approach is far more effective. When leaders begin by recognising the expertise that already exists in a staffroom, improvement feels less like correction and more like refinement. Teachers are far more willing to explore new techniques when those techniques are positioned as strengthening the craft they’ve already developed. In other words, respecting professional expertise doesn’t weaken a reform. It creates the conditions for it to actually work. This part is tricky because it relies on communication and structures that recognise expertise, while still guiding staff in a direction that sits in alignment with schoolwide instructional practices.

While I would love to just show people the research on an instructional technique, give some examples and some opportunities to rehearse, this doesn’t necessarily land well. I’d love for those who are experts to take any new learning from that or chime in with extra insights, but that model of presenting a focus technique doesn’t really allow for that. Instead, it's typically far better to start by identifying where strong practice already exists. Before introducing a new instructional technique, look for classrooms where elements of it are already happening. Highlight them. Invite those teachers to share how it works in their context. Position the practice as something that already lives within the staffroom rather than something imported from outside. Importantly, don’t invite critique.

This does two things. First, it acknowledges the expertise that already exists among teachers. Second, it allows new practices to be introduced as refinements of familiar routines rather than replacements for them.

Similar to before, teachers should also use their expertise to help shape how that practice looks in their different subject areas and year levels. It’s about considering how a technique looks in different contexts and engaging in conversations about these. Those conversations allow the framework to develop through professional problem-solving rather than passive adoption.

In other words, improvement works best when teachers feel like contributors to the work, not recipients of it.

But if your goal is to generate maximum resistance as quickly as possible, presenting the new pedagogy as the long-overdue fix for everything teachers have been doing wrong remains a very reliable strategy.

5) find a new shiny thing

Just like fashion trends, ditch the old and bring in the new when the media (the most reliable source of education truth) or your favourite online teaching forum starts promoting the next new shiny thing. You’ve seen 21st century skills. You’ve seen discovery learning repackaged under about a hundred new names. You’re looking at science of learning now. But wait until AI-integrated multi-dimensional learning systems9 comes along! You obviously should throw out all the gold leaf and great instructional work your teams have done and start again. Change fatigue is a myth anyway.

I’ve said it before and I’ll say it a hundred times over. Professional learning should be iterative.

The goal of professional learning isn’t to constantly replace one instructional idea with another. The goal is to refine practice over time. The most successful schools I’ve seen don’t lurch from framework to framework every few years. They identify a small number of high-leverage instructional practices and spend years strengthening them.

That means revisiting them. Refining them. Getting better at them.

Research on improvement consistently shows that sustainable change rarely comes from dramatic resets. It comes from disciplined attention to a small number of practices over a long period of time. Meaningful improvement happens through continuous cycles of experimenting, learning and refining10. Notice I didn’t say repeatedly throwing everything out and starting from scratch.

If you’re serious about changing teaching, resist the temptation to chase the next shiny thing.

But if your goal is to guarantee change fatigue and ensure that no initiative survives long enough to make a difference, constantly replacing your pedagogical framework is a remarkably reliable strategy. Staff will really love it.

the first five steps: sabotaging a schoolwide pedagogical shift

If your goal is to ensure your brand new pedagogical framework quietly fades into the long list of well-intentioned initiatives that never quite made it beyond the launch event, the five strategies above should serve you extremely well. Feel free to layer on your own as well. It could be a lot of fun, if only our literal students’ futures weren’t collateral in our failed attempt at organisational change.

It sounds ridiculous to write it out like that, but versions of these techniques appear time and time again in schools attempting large-scale pedagogical shifts. I’ve been actively involved in this space for a while now, and have seen things go terribly often. Anyone involved in this area will tell you the challenge with shifting pedagogy isn’t the pedagogy. It’s the people. This isn’t just a competition about who knows the most about cognitive science. This is a game in organisational change. And the research on organisational change consistently suggests that large-scale improvement tends to happen slowly, through repetition, refinement, and professional learning over time.

In other words, the real work of improving teaching rarely looks like a dramatic launch. It looks much more like teachers experimenting with small changes, refining techniques together, and gradually building habits that become embedded in classrooms.

Now that is considerably less glamorous than a launch event. But it does tend to work a bit better.

So if the five steps above are about how to guarantee spectacular failure, the next question is obvious: what would the first five steps to success look like instead?

As it happens11, I’ll be unpacking this exact topic at researchED Ballarat on Saturday for those of you lucky enough to have secured tickets before they sold out. Drawing on research on organisational change and implementation, I’ll be outlining a few practical steps leaders can take when they’re thinking about how to introduce and embed large-scale instructional change across a school.

Because while it turns out there are many ways to ruin a pedagogical shift, there are also some surprisingly consistent patterns behind the ones that actually work.

Share

“What makes you think that what I do doesn’t work?”

“What makes you think you can’t be better?”

I would describe myself as an extrovert who barely knows what the inside of her own house looks like. If there’s an invitation, I’m either there or I’m the one who sent it.

Still being only 8-ish months into living in Queensland, I’ve been lucky enough to be in a lot of situations where I’ve met new people, which is a thing I love doing. What I love slightly less is talking about my line of work with other teachers.

But, Caiti, aren’t you counting down the days until researchED Ballarat just so you can talk about your work with other educators??

You’re totally right, I am. I genuinely cannot wait to talk shop with educators who have actively chosen to be in that room. That’s different. At conferences, people opt in. In the wild, they don’t. If science of learning sneaks its way into a social conversation with teachers, it doesn’t always land neutrally. Teachers have had various experiences of evidence-informed practice that inform their viewpoints on this topic.

So sometimes, in social settings, I skim over the details of what I do. Not because I lack conviction - as you’ll know if you’ve read my posts - but because I’m aware that not every dinner table needs to turn into a pedagogical standoff1.

Recently though, the skimming didn’t happen.

A friend - perhaps generously, perhaps misguidedly - mentioned to their teacher friends that I do some writing and speaking in the education space. I suspect it was intended to be a point of interest. Instead, it functioned more like a starter pistol.

I cautiously explained that I write and speak about the science of learning or evidence-informed practice in efforts to ensure that students across the world have equitable access to high quality instruction.

And then it came.

“What makes you think that what I do doesn’t work?”

There it was. Not curiosity. Not interest. Not even disagreement.

Defence.

The interesting part is I don’t know how this person teaches, let alone whether or not it works. That wasn’t part of the conversation. There was no critique of practice. I didn’t speak to concerns I have about some common practices in education. I just named the space I work in.

It didn’t matter. The assumption had already formed: if I spend my time talking about using evidence to inform and improve practice, I must be operating from the belief that other people’s practice is deficient.

So, I replied.

“What makes you think you can’t be better?”

the song beneath the words

I’ll spare you the details of the rest of that relatively short-lived exchange. The conversation over dinner moved on. But that line lingered for me.

I sometimes joke that I’m like the Taylor Swift of science of learning. If you say something that doesn’t sit right with me, I’ll write about it. And this one didn’t sit right with me. Because beneath that question wasn’t disagreement. It was protection.

In their book Leadership on the Line, Ronald Heifetz and Marty Linsky talk about listening for the ‘song beneath the words’. This is essentially considering the underlying subtext behind what’s being said. The song beneath the words reveals what feels at stake.

On the surface, the question was about effectiveness of instructional techniques:

“What makes you think that what I do doesn’t work?”

But the song beneath this sounds a little different. It sounds a bit more like:

Are you saying you’re better than me?
Are you saying that I’ve been getting this wrong?
Are you saying I’m not a good teacher?

That’s not a conversation about pedagogy. It’s a reflex to protect identity. And it’s not the only time I’ve heard similar commentary with a similar subtext, which is why I’m talking about it2.

Teaching is deeply personal work. Our classrooms aren’t merely office spaces that we clock into at 9am and out of at 3pm. They become extensions of us. Our routines, our explanations, our presentation styles - they’re tied up with who we believe ourselves to be as professionals.

Most teachers don’t just do teaching. They are teachers.

And I would argue that most people in this profession don’t just want to be competent. They want to be great. They have woven that aspiration into their identity. That is precisely why conversations about improving practice can feel loaded. If being a ‘great teacher’ is part of how I understand myself, then the suggestion that my practice could be better doesn’t land as neutral feedback. It lands as a challenge to that identity. And of course, when identity feels exposed, protection kicks in.

The irony here is that I hadn’t even had the chance in that conversation to evaluate or critique any of that teacher’s practice. I simply described a space centred on refinement. However, in order to address this reasonably common viewpoint, we need to understand the following.

Refinement carries an implication: that better is possible.

And if better is possible, then perhaps I’m not the finished version of the teacher I imagined myself to be.

That’s the part that can sting.

“it works”: a shield for complacency?

Here’s the professional tension.

Something can work and still be worth improving. Those two statements are not contradictory. Yet, in education, “it works” often functions as the end of the conversation rather than the beginning of it. But what does “it works” actually mean?

Does it mean that students are compliant? That they enjoy lessons? That they like their teacher? That student results are acceptable? Or that the instructional approach is as strong as it could be for as many learners as possible?

Perhaps that’s the core of the issue. “Working” isn’t a fixed benchmark; it’s a relative one. But relative to what? Who decides what counts as evidence that something works?

We can point to results. We can point to student progress data. We can point to student surveys. Or we can reference broader qualities like confidence or empathy - important outcomes, but ones that are difficult to measure and directly link to specific instructional choices.

At the end of the day, “it works” is a weakly defined descriptor. It sounds definitive, but it rarely is. And maybe that’s why in many other professions the conversation doesn’t stall at “Does it work?” The question is usually, “Can it be improved?”

Professional athletes refine technique even when they’re winning. Surgeons refine procedures even when operations are successful. Engineers revisit designs that already function. Iteration isn’t seen as an insult to competence, but as evidence of it.

This is where “it works” becomes problematic when it is used as a shield against improvement. This isn’t because the practice doesn’t necessarily ‘work’, but because the statement closes the door to refinement. It signals satisfaction with adequacy. It’s a shield for complacency. If the standard for professional practice is simply that something functions, then the bar is low. Very low. Our students deserve more than methods that merely function. They deserve practitioners who assume growth is ongoing, who treat craft as iterative, and who are willing to sacrifice comfort for the chance to be better.

To go back to a quote from a previous post of mine, stop adding and start improving:

To be clear, I don’t see “it works” as a shield for complacency. It’s a protective mechanism when a threat exists. And this isn’t just a threat to professionalism - it’s a threat to identity.

improvement without indictment

I’ve spent a lot of time buried in research around organisational change management for an extended period now. The patterns are remarkably consistent. When identity feels implicated, resistance isn’t irrational. It’s protective. To draw on Heifetz and Linsky again, resistance is often a response to a perceived loss - status, certainty, competence, or even identity.

(At this point, it would be remiss of me not to mention that I’ll be unpacking the process of shifting an entire schoolwide pedagogical framework in depth at researchED Ballarat on Saturday 14 March 2026, for those of you attending.)

So if identity is what’s at stake, the question becomes this:

How do we talk about improvement without making it feel like an attack?

Because the goal isn’t to win the argument. It’s to shift the culture.

As a disclaimer, I do have some strong viewpoints on this sentiment (see below from a note I posted here on Substack, inspired by this incredible post by David Didau):

Is teaching about identity or evidence? And if it’s about identity, whose? Because even if I were of the opinion it was a matter of identity, it wouldn’t be that of the teacher.

We can say that by neglecting effective academic instruction, we’re fostering students who are somehow holistically ‘better’, but who’s that conversation really about? The student? Or is it secretly about the teacher who doesn’t want to give up their personal instructional preferences?

Those are some confronting questions, but confrontation doesn’t change culture either. Instead, leaders need to disarm the threat if they truly want to see change happen. That begins with separating worth from work. Professional learning cannot be framed as a verdict on competence. It has to be framed as the ongoing development of craft. It’s the refinement of instruction. Practice must be positioned as improvable without implying inadequacy. Iteration has to be normalised - and modelled publicly.

In many cases, the most effective move is to understand the underlying concerns behind what looks like resistance, and address it directly. Teachers have built their practice over years, often with inconsistent preparation and probably sub-par support. They have figured things out through experience. They have developed strategies that do, in fact, work. That matters. And that should be acknowledged as part of any conversation around refinement.

Acknowledging existing expertise is not a concession. It is a prerequisite for change. If leaders dismiss or devalue - even inadvertently - what already exists, they reinforce the very threat that they are trying to disarm. When professional judgment is recognised and respected, refinement feels less like correction and more like collaboration.

This means sharing existing practice that is effective amongst staff, instead of purely focusing on what’s new. It means explicitly positioning new learning as ways of approaching existing techniques differently, and being willing to admit when these are less effective than what was already being done. It’s about changing the conversation around professional learning from being centred on adopting something new to instead be permission to amplify already-effective practice. I often say to teachers that if it’s something the research says is likely to be effective, teachers will almost always already be doing it in some form or another.

Specifically, this means beginning by seeing how something is currently done before ever suggesting an alternative. It means asking, “where is this working well?” before asking how it could be made sharper. It means naming what is already strong in the classroom and using this as the springboard for refinement. Rather than allowing the threat to grow, it’s about using clarity to reduce it.

Most importantly, it means modelling the same vulnerability you’re asking for. Publicly revise your own practice. Admit when you tried a new technique and it didn’t land. Be willing to show that you’re learning too, and importantly, that you don’t have all the answers.

so what makes you think that what i do doesn’t work?

Let’s return to where we started.

“What makes you think that what I do doesn’t work?”

The honest answer is that I don’t. I don’t know how you teach. I don’t know your students. I don’t know your outcomes. That was never the claim.

What I don’t assume, however, is that “it works” is the end goal. The purpose of professional practice is not to defend what functions; it is to refine what we do in service of those we teach. Something can work and still be strengthened. Something can produce results and still leave room for greater clarity, precision, or impact.

The question shouldn’t be about whether it ‘works’. It should be about whether we are willing to keep improving it.

When I ask, “What makes you think you can’t be better?” I’m not questioning competence. I’m questioning finality. I’m questioning the underlying implication that we have arrived. In a profession as complex and consequential as teaching, arrival should make us far more nervous than iteration ever could.

The role of the leader in these scenarios is to disarm the concerns that close off conversations about improvement. It is to make refinement feel normal rather than threatening and collective rather than corrective. Leaders do not lower the standard to avoid discomfort; they lower the temperature so the standard can rise.

So no, I don’t assume it doesn’t work. I assume we’re not finished. And we shouldn’t want to be.

Share