S5 E16 - Is Whole Class Phonics Instruction Really Equitable?

Subscribe to the Podcast

Hi there, welcome to this episode of the Structured Literacy Podcast. It's Jocelyn here, coming to you from Tasmania, the home of the Palawa people. At Jocelyn Seamer Education, we believe that every child has the right to be taught with evidence-informed instruction and that every teacher has the right to be supported to make that happen.

Today, I want to tackle a question that comes up frequently in professional learning sessions and in various online places: Should phonics instruction be delivered to the whole class in your own classroom or in targeted groupings? This is a topic where we often hold strong views and where we are currently receiving very mixed messages, so let's explore this issue together.

Proponents of whole class phonics instruction often cite equity as their primary reason for the recommendation. The argument goes that all students deserve access to the same high quality instruction and that separating students into groups risks creating or reinforcing inequities, where some students receive less comprehensive instruction than others. And I would say, if we were talking about science or HASS or our text-based unit, that I would agree completely, but we're not. We are talking about an area of literacy learning that is constrained, that needs to be taught cumulative, with mastery developed of one thing before we move on to another. There's also the view that if we don't expose struggling students to the content their peers are learning, they'll fall further behind and never catch up. Many teachers worry that by placing students in what might be seen as lower groups, we're limiting their exposure to essential content and potentially creating a self-fulfilling prophecy where we don't expect enough from them. I'll get to this issue of exposure in a little bit.

Now these are important considerations, and when we have concerns like this and questions that come up, I want to encourage everyone to ask them. If we keep them to ourselves, it's likely that we never develop deeper understanding of how human brain architecture operates, and we are running on worries rather than on empirical evidence. So we want to be empirical in this discussion and not emotional. Every teacher wants to give their students the best possible chance to succeed, and we absolutely should be concerned about equity in our classrooms. However, when we examine this particular argument through the lens of cognitive science, and what we understand about how human brain architecture works, a bit of a different picture may emerge, and I think it does.

The primary reason for teaching phonics in targeted grouping centres on cognitive load and the need to support working memory. Research tells us that working memory is extremely limited. I, along with 450 other people, had the privilege of hearing John Sweller talk about cognitive load theory live at the Science of Learning conference on the Sunshine Coast last week. He reminded the audience that humans can really only process two to three novel or new bits of information at any one time, and he was hesitant about the three. He questioned himself on that in the moment and said actually, let's call it two. So keep the number two in mind, because we're going to do a little exercise here on the podcast in a moment.

In order to complete complex tasks, we need to draw on information held in long-term memory. When information has been learned, as in, it sits in long-term memory, we can introduce many, many elements into a task, and it's fine. The constraining effects of working memory no longer apply. But if we don't have information in long-term memory, our working memory is very easily overwhelmed.

So let's do an exercise. I want you to imagine a Year One student or a Year Five student or a Year Eight student who does not have the basics of phonics, including handwriting, stable in their long-term memory. And if you're thinking, well, Jocelyn, how often would we encounter a Year Five or a Year Eight student, I would say, more often than we're comfortable talking about. I've met them, not all of them, obviously. Some of them I've encountered along my journey.

Each time students write letters they are likely to do it different ways if they have this learning profile. And when they do write things you can see them actively processing the letter formation. They aren't sure of short vowels and that's a very common situation for older students with reading difficulty and they often mix up consonants like <ch> and <sh>, reading them interchangeably. They can't tell the difference between long and short vowels to listen to, as in they can't tell you that man has a short vowel and main has a long vowel, and there are a few letters in the basic code that they are slow to retrieve and often get mixed up, like G and J, not knowing whether it's <j> or <g>.

In the name of equity, let's say we are teaching a Year One phonics lesson, whole class. So a student with this learning profile is sitting in the lesson for the vowel digraph double e, as in the word "sheep", we introduce the double e, all is well. We ask the students to write the double e and this is where the first hurdle comes because this student does not know how to write the e automatically. Every time they write it they're almost starting in a different starting spot, they're retracing over their letters, they're missing bits, they're getting it backwards, so just the act of letter writing is effortful.

Now we have two elements that need to be processed in working memory and we are approaching the limits of the student's working memory capacity. But they're hanging in there because they want to please and they are being largely compliant. Then it's time to read some words and the first word is sheet. So we now have our third element that has to be processed through working memory, the digraph <sh>, and the student knows <t>, so that's ok. But we've already reached the limit of working memory for any human because the student isn't automatically writing e, the vowel diagraph double e is new and they aren't automatic on <sh>.

So your attention is now drawn to this student because you, as a teacher, know they need extra scaffolding. But you're also conscious that your fast lane learners have already finished writing sheet and they're now getting restless. You, as the teacher, are now torn between helping this student or group of students, which is not uncommon, who fit this profile, and stretching the more capable students, you cannot do both. The next word is is cheek. And here these students are in real trouble, because not only is double e new, the handwriting is not automatic and the student's still wobbly on <ch>, but the <k>, that's not automatic either, so this student's capacity to attend is now completely blown out of the water.

I find it very hard to see the equity in this situation for either the student who needs more support or the student who needs stretch, because neither student is getting what they need. And at the Science of Learning Conference on the Sunshine Coast, I had the opportunity to talk for about half an hour with a neuroscientist named Jared Cooney Horvath, and I posed this exact question to him and asked him whether I was on the right track here, because I'm not a scientist, but he is. So I asked him, am I right in my thinking about what this student needs to learn? And he said resoundingly yes. But let's come back to the idea of equity and catching up students who are behind their peers.

One of the suggestions made in a whole class teaching situation is that the student needing extra support then goes off to intervention to fill the gaps that exist in their code knowledge. Ok, that might seem to make sense, except that we're now asking the student with the leakiest memory, now that's my term, not an official cognitive science one, to learn twice as much content as their peers, and that? That doesn't make any sense. The student who needs the highest intensity instruction, who requires the most opportunity for repetition to encode learning into long-term memory, has to learn twice as much. This poor kid is being dragged through phonics content, achieving either no success or such a small amount of success that the adults around are asking, is it worth investing all of this time in students whose growth is so painfully slow and, at this rate, will never learn to read? Let's come back now to this idea of exposure.

There is nothing in cognitive load theory that says we will learn phonics by exposure. Yes, we learn tier one language, everyday language, through exposure, because we are hardwired to learn that. We are not hardwired to learn to read or to spell. We do not learn through exposure. We learn through active engagement, with repetition after repetition, including context such as words, that's why they're in the phonics lesson. But just showing us something and us being compliant in copying it is not the same as active engagement. Exposure doesn't lead to learning, but engagement does.

For students who require a higher intensity of instruction to move knowledge from working memory to long-term memory, being taught content they're not ready for, because phonics is taught cumulatively, doesn't lead to catching up. It often just leads to confusion, disengagement and poor behaviour. The decision about whole class versus targeting groupings isn't one driven by ideology. At the start of the episode I said we're going to be empirical here, not emotional, so we have to have our decisions be driven by assessment data and our understanding of the student's learning profiles. Now there's no test that tells you precisely how many repetitions a student will require to encode new learning into long-term memory, so we need to use some teacher judgement, based on what we know of the student, to make decisions on their behalf. Then we examine the data to see has this worked for them? If it hasn't, we need to give something else a go. And I don't mean something else as in a different style of teaching phonics. I simply mean we need a higher intensity here and we keep going until we find the level of intensity required. Now when I say level of intensity, I don't mean just conduct a faster phonics lesson. I mean smaller chunks of information with lots more consolidation time between the introduction of new bits of information so that the student doesn't just have a trail of gaps in the learning wake.

In my book Reading Success in the Early Primary Years, I discuss how we could approach differentiation in phonics instruction, and here's a quote.

The aim of any phonics instruction is to keep the largest number of students possible together in the main group of learners. It is much easier to teach a group of students who are all within a similar range because you can target content and pace of instruction to meet the student's needs. How close you get to this ideal will depend on your school context, the type of instruction that has occurred in the past and the opportunities you provide within your lessons for review and consolidation.

I have never, even in schools where students come from very privileged backgrounds, found a group of students where there weren't students with the needs I've described in this episode.

I'm not saying we have to go back to group rotations. In fact, I think it would be a mistake and a colossal waste of instructional time to do so. And I'm aware that there are programs that are structured in this way, one whole class lesson for everyone and then a good 40 to 50 minutes of group rotation. I'm not an advocate for that, unless it's 100% necessary in terms of the group rotation. I wouldn't be advocating for whole class phonics teaching, regardless of what we're seeing in student learning profile. What I'm advocating for is instruction that is targeted at student need and conducted in a way that responds best to our context and the students. Now some educators worry that students will feel less than if they end up in the lower group, and this is an understandable concern, I know where it comes from. But we're not calling the groups the wombats and the eagles. We're just saying it's time for you to go for your reading lesson now. Off you go with Mrs Smith.

The alternative is students who are constantly presented with material they cannot access, students who experience repeated failure because the content is beyond their capabilities and they're always just getting a finger hold on new things before the class moves on. Students like to feel successful. Targeted instruction helps them experience success because we're actively supporting their working memory and managing their cognitive load. Success breeds confidence, and confidence leads to greater effort and engagement. One of the things that schools consistently tell me is that when they adopt a targeted approach to phonics instruction, text-based unit work is done whole class back with your own teacher in your own classroom. But when they adopt a targeted approach for phonics instruction, the instances of negative behaviours drop dramatically. And that's because students can either be on task and learning or off task, because either they already know it or they can't access it. Where the grown-ups have responsibilities here to engineer a learning environment that helps everybody be engaged in active learning for as many minutes as humanly possible.

At the other end of the spectrum from the student with the leaky memory are the students who already have the code knowledge being taught, lessons for them are essentially a waste of time. If they can recognise graphemes, recall them without effort and write them down, if they can read them in words, if they can write them in words, if they can read them in text and can use them, where's the learning for them? So it's about asking, what are students getting out of this lesson and how much of the lesson is actually providing desirable difficulty for those students? If it's four minutes of a 40-minute lesson, well, I don't think that's quite good enough. Managing cognitive load isn't just about preventing it from being too high. We're looking to optimise intrinsic load, which means that it's also not too low.

If there's nothing here for us to learn, we will disengage. Advanced students need appropriate challenge just as much as struggling students need appropriate support, and there are nuances in this and you can at times provide that stretch within the main lesson. So I'm not saying you can't, but we have to be realistic about where the students are. One of the ways you can use data to determine whether we're in the zone here is to have a look at how many new graphemes they are learning to automaticity per term. If we take the whole of the early years as one progression and students are actively learning to automaticity to recognise eight to ten graphemes a term, then they're learning. If they're not, something needs to happen. Another little quote from Reading Success in the Early Primary Years:

Different children will come to this understanding with differing levels of ease and speed. 

And that's from Stanislas Dehaene's book How We Learn.

Some children, such as those with dyslexia, are more vulnerable to cognitive overload than others.

So in this way, while the core methodology is the same, the pedagogy is consistent, the starting point and the pace of instruction is not one size fits all. So how do we move forward in what feels like a really messy area? Well, the answer is, it depends, and I'm sorry that I don't have one consistent answer that can be applied to every setting, because the answers here depend on your students, your context and what your assessment data tells you about student need.

So some questions to consider when making this decision are:

1. What does my assessment data tell me about the range of phonics knowledge in my class? That's the first thing. What does it tell us about the needs of students and where they're going next in learning, remembering that at the start of your journey into structured literacy, your data, hopefully, will look very different than it does three years in.

2. Can I reasonably meet the needs of all of my students within a whole class setting through differentiation? And the answer here is all wrapped up in how many pieces of novel information will my students have to process to participate in this lesson? If it's more than two, it's too many. If it's none, there's not enough.

Now let's think about the student who there's more than two bits of novel information and some of the circumstances where we may be able to make adjustments to support students to participate rather than grouping them out. So, for example, if a student has dysgraphia, they can use magnetic letters or letter tiles to participate in the word building part of the lesson. We know from our data with that student that they can recognise graphemes. They just really struggle to write them. If we give them five letter tiles and we say a phoneme, they can automatically point to the correct grapheme. So we know they've got the knowledge, just the writing that's really hard. So the adjustment here is what they can do instead of writing by hand. You may also have students who find speaking almost impossible for a whole variety of reasons, but they can actually recall and write graphemes when we say the phonemes automatically. They can write words when we say a word and we give them four or five words, they can with ease find the correct word. So there are ways that we can provide adjustment within the lesson for them to demonstrate their knowledge. These are true adjustments. These are equitable. This is what we need.

3. What resources as in time, staff and materials do I have available to support differentiation? Am I working in a two-teacher school where I have three or four grades in the one classroom? I've done that. In that case, rotations of some sort are unavoidable. Or do we have four Year One classrooms and four reasonably experienced classroom assistants? In this case, we have eight adults who can take a phonics lesson, so we can provide incredibly targeted instruction at the point of need for the students. I'm not saying it will always be guaranteed that all students have to move. Maybe a learning support teacher takes the students needing the highest intensity of instruction, because everybody else is in the ballpark and can be supported in the main lesson.

This thinking about student need and how we best use the resources we have to meet that need, as we consider cognitive load, is a much more equitable way to go about solving this issue. Because these students won't then be getting 40 minutes a week of targeted instruction, they'll be getting 40 minutes at least of targeted instruction per day.

Ultimately, the most important factor isn't whether we're ticking a box on teaching whole class or in targeted groupings. It's whether our instruction is responsive to student needs and based on assessment data. And how do we know we've got the formula right for our students? When every single student is growing at the optimal rate for most students, as in those who do not have cognitive disability or other significant processing challenges. The "at least" that we're looking for is eight to ten graphemes per term, learned to automaticity, to recognise and to write the grapheme, so you say, write the /a/ and they can write the /a/, not necessarily to use within their own independent writing accurately all the time. It's your data that gives you the feedback. Now your DIBELS data won't give you that. It's not what it's designed for. You need to be doing phonics focused check-ins to determine whether the learning that we have aimed for has stuck.

The debate around whole class versus targeted phonics instruction has become a little polarised, but like most things in education, the reality is nuanced. What works in one context may not work exactly the same way in another. The core of phonics instruction, however, is context independent. It is not that in some places we're teaching explicitly and in other places we're making play-doh letters. Everyone deserves the same core strong, robust teaching, just targeted at the point of need and done in a way that responds to what we know about their processing.

Instead of adopting a rigid stance, I encourage teachers, schools and leaders to be flexible and responsive, to use assessment data to guide their decisions and to focus on the ultimate goal, which is helping every student develop the skills they need to become confident and proficient readers and spellers. And will we get it right all the time? Of course not. One of the other things I learned at the Science of Learning Conference is that creativity and problem solving requires three things: it requires knowledge, it requires time and it requires the opportunity to fail. But just remember, we don't want to fail big for our students. Let's just fail little so we can make adjustments along the way. And there's a fourth element here that Jared Cooney Horvath suggested really should be in that criteria for problem solving and creativity, and that is structure. So we need to know what structures we're using to help everybody get the learning we know they deserve. Until next time, everybody, happy teaching, bye.

Looking for phonics resources to support student learning? Join us inside The Resource Room!