I Do, We Do, You Do: Gradual Release of Responsibility

I do, we do, you do is easy to write on a lesson plan and hard to run well. The decision that matters is when to hand over: model for too long and the class sits watching, release too early and half of them practise errors. This guide uses the research on worked examples and guided practice to say what each phase asks of you, and how to tell, during the lesson, that the class is ready for the next one.

‍

The approach is often linked to Pearson and Gallagher's work on reading comprehension. It is also linked to Vygotsky's account of learning through guided social interaction (Pearson & Gallagher, 1983; Vygotsky, 1978).

The value of the model is not the slogan itself. It is the careful shift in cognitive responsibility. Evidence from modelling, guided practice, and retrieval practice suggests that learners need clear examples and repeated independent recall (Rosenshine, 2012; Karpicke & Roediger, 2008).

‍

What is 'I Do, We Do, You Do'?

I Do, We Do, You Do supports structured teaching. It works by slowly passing responsibility from the teacher to the learner.

First, the teacher shows the skill (I Do). Next, the teacher and learners practise together (We Do). Finally, learners use the skill independently (You Do).

Vygotsky's work and Pearson and Gallagher's 1983 research underpin this model. Support decreases as learner competence grows.

The "I do, we do, you do" sequence links to established teaching ideas: teacher modelling, guided practice, scaffolding and gradual release. It is better treated as a structured lesson routine than as a programme with a single guaranteed success rate.

‍

Where it comes from: the gradual release of responsibility model

The gradual release of responsibility model was proposed by Pearson and Gallagher (1983) in a review of how reading comprehension was being taught. Their diagram runs from a task where the teacher holds all the responsibility, through joint responsibility, to a task where the learner holds all of it. It was a framework for teaching comprehension strategies, not a classroom trial, and it never mentions three fixed slots in a lesson. Duke and Pearson (2009) restated it in the same terms: the point is the transfer of responsibility, however many steps it takes.

Fisher and Frey (2008) turned the idea into four phases in Better Learning Through Structured Teaching: focused instruction, guided instruction, collaborative learning and independent learning. Their later paper sets out the same four phases and credits Pearson and Gallagher for the model (Fisher & Frey, 2013). The fourth phase, learners working together before they work alone, is the one most three-part versions leave out. Webb, Massey, Goggans and Flajole (2019) reviewed thirty-five years of the model and argued that it was always meant as a responsive scaffold, adjusted lesson by lesson, rather than a template.

The words "I do, we do, you do" come from a different tradition. Archer and Hughes (2011) put "I do it, we do it, you do it" at the centre of Explicit Instruction, alongside a review of prior skills before teaching, frequent responses from every learner and prompts that fade. Rosenshine (2012) never uses the phrase. He describes modelling, guided practice until learners reach a high success rate, and independent practice, which is the same shape with different names. So the slogan is a summary of explicit instruction, and the gradual release model is the older account of why the hand-over matters.

That history explains a search habit. Teachers in the United States say "gradual release of responsibility" and "gradual release model"; teachers in the UK say "I do, we do, you do". They are asking about the same practice. The difference worth keeping is this: gradual release is about who is holding the responsibility for the thinking at each moment, and a three-box lesson plan can hide that question rather than answer it.

‍

Why I do, we do, you do works: modelling, worked examples and cognitive load

The model works because of what it does to working memory. A novice who is handed a new problem and told to solve it has to hold the goal, the steps and the content at once. Sweller (1988) showed that this kind of unguided problem solving overloads working memory and leaves little capacity for learning the method. Kirschner, Sweller and Clark (2006) made the same case across discovery and inquiry approaches: with novices, minimal guidance fails.

The "I do" phase is a worked example with the thinking said out loud. Sweller and Cooper (1985) found that algebra learners who studied worked solutions learnt more, in less time, than learners who solved matched problems. Chen, Retnowati, Chan and Kalyuga (2023) repeated the finding on multi-step maths, with better retention and transfer and lower load at each step. The example does the heavy lifting; the narration shows learners which decision comes next and why.

Thinking aloud is the part teachers most often skip, and it is the part with an outcome attached. Fisher, Frey and Lapp (2011) coached middle-school teachers to think aloud during modelling. Their practice changed, and their learners' reading scores rose. The strategy instruction that the "I do" phase models also lasts: a meta-analysis by de Boer, Donker, Kostons and van der Werf (2018) found an effect of g = 0.50 at post-test that rose to 0.63 at follow-up, with the largest gains for learners from low-income homes.

Two honest limits. Stockard, Wood, Coughlin and Rasplica Khoury (2018) reviewed 328 studies of Direct Instruction curricula and found every estimated effect positive, but Direct Instruction with capital letters is a scripted programme, not the three-part lesson. And on delayed transfer, Likourezos and Kalyuga (2017) found that examples-first and problem-solving-first came out the same. The order is not magic. What matters is the guidance a novice gets while the thinking is still hard.

‍

‍

‍

The Three Phases Explained

Teachers show learners skills through modelling ("I Do"). Teachers and learners work together when scaffolding ("We Do"). Learners build confidence through independent practice ("You Do"). This structure gives responsibility to the learner.

‍

Open the full-size image

‍

I do: the modelling stage

Effective teaching starts with explicit instruction. The teacher shows the new skill, breaking it down into small steps (Rosenshine, 2012). This helps every learner to understand and practise successfully (Archer & Hughes, 2011).

Some teachers run a first pass in silence, the 'silent teacher' approach, modelling each step of the new skill without narration so learners watch the hands and the board. That can lower the load on a first look, but it is not the modelling the evidence measures.

The narrated pass is the one that counts. Once the silent pass is over, work the example again and say each decision out loud: what you are looking at, what you do next and why. Coaching teachers to do exactly this raised their learners' reading scores (Fisher, Frey and Lapp, 2011).

‍

We do: the guided practice stage

Learners get support to find the right answers via collaboration (Vygotsky, 1978). Teachers give less guidance as learners tackle three to five questions. Each question breaks down into steps for learners (Wood et al., 1976).

Interactive tasks ask all learners to answer questions. Effective questioning involves every learner and shows whether they understand (Rosenshine, 2012). This helps teachers check learner comprehension quickly.

‍

You do: the independent practice stage

This is the time for learners to put into practice what they have learnt during the first two stages by practising the new skills independently. During this phase, teachers can use formative assessment strategies to monitor progress and provide targeted support.

Learners ask questions, and teachers offer support. Most learners should work alone on tasks (Vygotsky, 1978). This allows flexible pacing and tailored help.

‍

‍

‍

What it is not: "I do, I do, I do" and the guided-practice confusion

The live argument among UK teachers is not about whether to model. It is about what happens after the modelling. Sherrington (2025) traced the confusion to the phrase itself: "we do" gets read as the teacher working a second example on the board with the class watching, which is a second "I do". Mark Enser, writing on Substack in April 2026, called the resulting lesson "I do, I do, I do": the teacher models, models again with some hands up, then sets the task, and the thinking never leaves the front of the room.

The research gives that objection a precise shape. In the studies behind the worked-example effect, the example was never the whole treatment. Sweller and Cooper (1985) paired each worked example with a problem the learner then solved. Renkl and Atkinson (2003) went further and faded the examples, removing one solution step at a time so the learner completed more of each problem than the last.

Schwonke, Renkl, Krieg, Wittwer, Aleven and Salden (2009) showed that faded examples beat a tutoring system on time and on conceptual understanding, so the effect is not an artefact of a weak comparison. Miller-Cotto and Medrano (2026) randomised 114 sixth graders across four conditions and found that fading gave the largest gain. That is what "we do" means: the learner does more of the work each time, with the support coming away in steps.

Guided practice in Rosenshine (2012) is the same thing described from the teacher's side. The teacher asks many questions, every learner answers, and the class stays in guided practice until success is running at about 80%. Archer and Hughes (2011) add two conditions that most three-box lesson plans miss: check the prerequisite skills before the first "I do", and get a response from every learner during it, by choral answer, mini whiteboard or a partner explanation. A silent demonstration followed by a set of independent questions is the failure both authors are describing.

Three questions tell you which lesson you taught. Did every learner produce an answer during "we do", or did a few hands carry the class? Did the support reduce from one example to the next, or did you work three full examples in a row? Did you have evidence of roughly 80% success before you said "you do"? A no to any of these means the lesson was "I do, I do, I do", and the fix is not more modelling but a completion problem, a prompt sheet and a whole-class check.

‍

‍

How to Implement Successfully

Teachers need strong subject knowledge to use 'I do, we do, you do' well and meet aims. Understand typical learner errors and plan lessons carefully. Think about adapting tasks to support all learners in the three stages.

Plan and time carefully for successful delivery. In the 'I Do' phase, think aloud to show learners your thinking (Fisher & Frey, 2013). When solving maths, verbalise each step and say why it comes next.

Learners show understanding through body language, questions, and answers; then move to 'We Do'. Circulate during group work; give feedback and adapt support. Some learners need more 'I Do'; others are ready for solo practice.

Documentation helps refine teaching. Note concepts needing more modelling and learners needing support. This data helps adjust lesson pacing. You can then spot learners needing differentiated help.

‍

‍

Advantages and Limitations

The model suits step-by-step skills such as a written method in maths or a practical routine in science. Teachers find it helps mixed-ability classes, because the hand-over can happen at a different moment for each group.

‍

The model has limits that teachers should note. It suits procedures with clear steps, and direct instruction of that kind is where the evidence is strongest. Once learners know the procedure, the same support starts to hold them back (Kalyuga et al., 2003). And for a task where the goal is to notice the structure of a problem before the method is named, a short, bounded struggle can come first (Kapur, 2008; Kalyuga & Singh, 2016).

Consider if content suits learners and if they're ready. Use this model for basic skills, but vary methods for complex thinking. Watch learners during practice. Adjust how you hand over control based on their confidence, not a schedule (Rosenshine, 2012; Fisher & Frey, 2013).

‍

Fisher and Frey's Four-Phase Gradual Release Model

The four phases are set out in "Where it comes from" above. Two things from Fisher and Frey (2008) are worth keeping beside them. The collaborative phase, learners explaining the method to each other before they use it alone, is the step most three-part lessons skip; a Year 8 class might go from modelling to shared notes to group writing before anyone writes solo. And the shift between phases is made on evidence, never on the clock: a piece of shared writing or a whiteboard round tells you whether the class is ready (Fisher & Frey, 2013).

‍

Rosenshine's Principles of Instruction and the GRR Model

Rosenshine's Principle 6 says teachers should model clearly. They show learners how they think and explain each step (Rosenshine, 2012). This "I Do" phase makes expert thought processes visible.

Principle 7 focuses on checking learner understanding. Teachers ask questions during modelling to see if learners grasp concepts. They use mini whiteboards to check readiness before "We Do".

Principle 8 aims for high success. Rosenshine (2012) advises guided practice until learners reach 80% accuracy. This helps ensure they are ready for independent work.

Rosenshine (2012) warns against massed practice. This means heavy repetition without spaced review, which can harm learning. He suggests mixing practice and review.

The "You Do" phase should recur across lessons, not be a one-off. Teachers should add short review sessions to prevent forgetting. For titration, use "You Do" across three lessons, plus recap and exit tickets.

‍

Formative Assessment at Phase Transitions

Knowing when to change phases is vital in GRR. Black and Wiliam (1998) showed that teaching improves when teachers collect evidence of understanding and act on it. Collect evidence at each I Do, We Do, You Do stage. Without it, you may move on too soon, leaving some learners behind.

Simultaneous response suits "I Do to We Do" well. Mini whiteboards let every learner show answers at once. Paired think-alouds ask learners to explain their thinking to a partner while you circulate and note who is stuck. Targeted cold-calling (random names) works better than asking for volunteers.

These methods show the class's understanding, not just the understanding of a few learners. For example, a maths teacher can use an error-filled column subtraction. Learners find and correct the errors before practice. This checks understanding and starts discussion about misconceptions.

Exit tickets check understanding after guided work. Black and Wiliam (1998) said short tasks show which learners need more help. Teachers use a question and self-rating to group learners. This informs who needs support, scaffolding, or more instruction.

‍

When to fade the scaffold: the expertise reversal effect

Support that helps a novice starts to cost a learner who knows the material. Kalyuga, Chandler and Sweller (2001) trained trade apprentices with worked examples and found the expected advantage, then watched it disappear after two sessions as the apprentices gained knowledge. Kalyuga, Ayres, Chandler and Sweller (2003) named this the expertise reversal effect. For a teacher it settles a question the three-box plan cannot: staying in "we do" past the point of readiness is not safe, it is a cost.

The signal to fade is success rate, not the clock. Rosenshine (2012) puts guided practice at about 80% success before independent work. Below that, learners are practising errors; well above it, the prompts are doing work the learners could do themselves. Mini whiteboards, a show of answers and a quick check of three learners' books give you the number in under a minute.

What about productive failure, where learners struggle before they are taught? Kapur (2008) found that a struggle-first sequence could beat instruction-first on later transfer. Kalyuga and Singh (2016) reconciled the two by asking what the activity is for. If the goal is to learn a procedure, examples come first; if the goal is to notice the structure of a problem before the method is named, a short, bounded struggle can come first, and the "I do" that follows lands better. Either way the release still has to be earned.

When the release is real, the learners who gain most are the ones who started furthest behind. Wu et al. (2023) ran a crossover trial of strategy instruction through gradual release with 1,688 learners in 30 classes. Comprehension and autonomous reading motivation both rose, and the weakest comprehenders gained most. The EEF guidance report on metacognition makes the same recommendations in plain terms: model your own thinking, then guide practice with support gradually withdrawn as learners become proficient (EEF, 2018).

‍

Implementation Mistakes to Avoid

The most common error is rushing the stages. Teachers feel pressured to quickly reach 'You Do', especially with curriculum demands. Moving too fast leaves learners without a strong base, because the support has come down before they can manage the task alone (Wood et al., 1976).

Some skills need the modelling spread over two lessons. A microscope routine is one: focus, slide, stage, light, then the drawing. Extend the 'I Do' phase rather than rush it, and check that learners can name the steps before they work in pairs.

Inconsistent support confuses learners during "We Do". Teachers sometimes give too much help, then too little. This affects learning.

This approach needs specific reduction points. When teaching essays, co-construct whole paragraphs first. Then, use sentence starters for learners. Finally, give key vocabulary prompts.

Teachers often see 'We Do' as just whole-class work. This misses opportunities with different group sizes. Fisher and Frey (2013) place a collaborative phase between guided and independent work for this reason: learners consolidate a method by using it with peers before they use it alone.

Rotate learners between whole-class practice, groups, and pairs in the 'We Do' phase. For column addition, teachers can solve problems as a class first. Then, teachers should offer support to small groups based on what they see.

‍

‍

‍

Subject-Specific Examples and Applications

'I Do, We Do, You Do' works well across subjects. Learners recognise its consistent structure. This familiarity reduces cognitive load. Learners then focus on the content (Fisher and Frey, 2013).

Writing models change how teachers approach English teaching. Teachers show paragraph skills and explain their choices as they write.

‍

Learners then build paragraphs together, suggesting words with teacher help (Vygotsky, 1978). After that, learners write alone, using techniques they saw demonstrated (Bruner, 1966).

This helps struggling learners who need clear examples (Graham & Perin, 2007).

The "I Do, We Do, You Do" model helps science teachers with practicals. Teachers first show learners microscope handling (I Do), following Wood et al. (1976). Next, learners practise with guidance (We Do). Learners then do independent tasks (You Do), ensuring skill mastery.

Bruner (1966) argued that an idea can be shown in actions, in pictures and in symbols, so column addition is modelled with blocks, then drawings, then digits. Learners solve problems together on whiteboards; Vygotsky (1978) noted fast feedback is key. Wood et al. (1976) found varied practice helps learners master the method.

In history, model how an expert reads a source first: Wineburg (1991) showed that historians check who wrote a document, when and why before they read a word of it, and that novices do not. Then analyse two documents together, with the class calling the checks, before learners work on a third alone.

‍

‍

Plan the release: a four-step scaffolding ladder

Plan the fade before the lesson, not during it. The research on faded worked examples gives you a ladder with four rungs, and the job is to decide which rung each part of the lesson sits on. Wood, Bruner and Ross (1976) called the support a scaffold because it comes down piece by piece; Renkl and Atkinson (2003) showed that the pieces should come down one solution step at a time.

1. Full model. You complete the whole task while thinking aloud. Learners watch and follow; a mini whiteboard question at the end checks they followed.
2. Completion problem. You complete every step but the last. Learners finish it, and you check the answer with the whole class before moving on.
3. Prompted practice. Learners complete most of the steps with a prompt sheet, a sentence starter or a partly filled grid. You circulate and take the prompt away from anyone who no longer needs it.
4. Independent practice. Learners complete matched problems alone. You only move the class here once guided practice is running at about 80% success (Rosenshine, 2012).

A Year 7 ratio lesson on the ladder looks like this. Rung one: you share 24 sweets in the ratio 3:1 aloud, naming the three steps. Rung two: you set up 40 sweets in the ratio 3:2 and stop before the final multiplication.

Rung three: learners do 60 in the ratio 2:3 with the three steps printed beside the question. Rung four: six problems with no prompts. The prompt sheet is the scaffold; the ladder says when it comes away.

Open the full-size image

The sketchnote above puts the whole sequence on one page: the three phases, the check before each hand-over and the fade from full model to independent work. Print it for a planning meeting, or keep it beside you while you decide which rung tomorrow's lesson starts on.

‍

‍

Free Resource Pack

The free download on this page is a ten-slide deck, I Do, We Do, You Do: Gradual Release of Responsibility, built as image slides for a department meeting or a fifteen-minute CPD slot: the blank-stare curve, Fisher and Frey's four phases, the responsibility matrix, the formative checks at each hand-over, scaffolding set against spoon-feeding, and a three-question lesson audit to run on your own next lesson. Enter an email address in the card below and the link is sent straight away.

‍

‍

Research by Rosenshine (2012) shows that explicit teaching of this kind is effective. Kirschner, Sweller and Clark (2006) explain why with cognitive load theory. Stockard et al. (2018) reviewed 328 studies of Direct Instruction curricula and found every estimated effect positive, with the caution that Direct Instruction is a scripted programme rather than the three-part lesson itself. The slides put these findings beside the three phases.

‍

‍

Frequently asked questions

‍

Who created the I do, we do, you do model?

Nobody created it under that name. The gradual release of responsibility model was proposed by Pearson and Gallagher (1983) in a review of reading comprehension teaching, and Fisher and Frey (2008) developed it into four phases. The phrase "I do it, we do it, you do it" comes from Archer and Hughes (2011) and their account of explicit instruction. Rosenshine (2012) describes the same sequence without using the slogan.

‍

Is I do, we do, you do the same as gradual release of responsibility?

They describe the same practice from two directions. Gradual release of responsibility is the model: responsibility for the thinking moves from the teacher to the learner over time. I do, we do, you do is a three-word summary of one way to run it. The risk with the summary is that it turns into three fixed slots in a lesson, while the model is about who holds the responsibility at each moment, however many steps that takes (Webb et al., 2019).

‍

What are the four phases of Fisher and Frey's gradual release model?

Focused instruction, where the teacher models and sets the purpose; guided instruction, where the teacher prompts and questions while learners try the task; collaborative learning, where learners work on the task with peers; and independent learning (Fisher & Frey, 2013). The third phase is the one most three-part versions drop.

‍

Is I do, we do, you do a form of scaffolding?

Yes, when it is done as described. Wood, Bruner and Ross (1976) coined the term scaffolding for support that is put up while a learner cannot manage a task alone and taken down as they can. The "we do" phase is the scaffold, and the faded worked examples studied by Renkl and Atkinson (2003) are the clearest research version of it. Three teacher demonstrations in a row are not scaffolding, because nothing comes down.

‍

How does I do, we do, you do fit Rosenshine's principles?

Rosenshine (2012) sets out modelling with worked examples, guided practice with many questions and checks, a high success rate of about 80% before independent practice, and independent practice with review. Read in order, those principles are the three phases with a condition attached to each hand-over. The condition is the part a three-box lesson plan tends to lose.

‍

How long should the I do phase last?

As long as it takes to model one complete example while thinking aloud, with a check that learners followed it. For most secondary lessons that is five to ten minutes. If learners need a second full example, the evidence on fading suggests making it a completion problem instead, so learners do the last step (Renkl & Atkinson, 2003). A skill with many steps can spread the modelling over two lessons rather than rushing it.

‍

Does I do, we do, you do work in every subject?

It works for anything with a procedure that can be modelled: a column method, a paragraph structure, a practical technique, the reading of a historical source. It suits open tasks less well, and once learners know a procedure the support can start to hold them back (Kalyuga et al., 2003). Use it to teach the method, then vary the task.

‍

What is 'I do, you watch; I do, you help; you do, I help; you do, I watch'?

A four-step version of the same sequence, often used in coaching and in special education. It splits "we do" into two steps: first the teacher leads and the learner helps, then the learner leads and the teacher helps. It makes the hand-over visible, which is its whole value. It carries no separate research base of its own; the evidence is the same worked-example and scaffolding research described above.

Plan one lesson this week around a single hand-over. Pick a procedure your class half knows, write the four rungs of the ladder on a sticky note, and decide before the lesson what evidence will move the class from "we do" to "you do". Then teach it, count the whiteboards, and notice whether the thinking left the front of the room.

‍

References

Archer, A. L. and Hughes, C. A. (2011). Explicit instruction: Effective and efficient teaching. New York: Guilford Press. ISBN 978-1-60918-041-6.

Black, P. and Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy and Practice, 5(1), 7 to 74.

Bruner, J. S. (1966). Toward a theory of instruction. Cambridge, MA: Harvard University Press.

Chen, O., Retnowati, E., Chan, B. K. Y. and Kalyuga, S. (2023). The effect of worked examples on learning solution steps and knowledge transfer. Educational Psychology, 43(8), 914 to 928.

de Boer, H., Donker, A. S., Kostons, D. D. N. M. and van der Werf, G. P. C. (2018). Long-term effects of metacognitive strategy instruction on student academic performance: A meta-analysis. Educational Research Review, 24, 98 to 115.

Duke, N. K. and Pearson, P. D. (2009). Effective practices for developing reading comprehension. Journal of Education, 189(1 to 2), 107 to 122.

Education Endowment Foundation (EEF). (2018). Metacognition and self-regulated learning: Guidance report. London: EEF. Updated 2021.

Fisher, D. and Frey, N. (2008). Better learning through structured teaching: A framework for the gradual release of responsibility. Alexandria, VA: ASCD. ISBN 978-1-4166-0635-2.

Fisher, D. and Frey, N. (2013). Gradual release of responsibility instructional framework. IRA E-ssentials, 1 to 8. Newark, DE: International Reading Association.

Fisher, D., Frey, N. and Lapp, D. (2011). Coaching middle-level teachers to think aloud improves comprehension instruction and student reading achievement. The Teacher Educator, 46(3), 231 to 243.

Graham, S. and Perin, D. (2007). A meta-analysis of writing instruction for adolescent students. Journal of Educational Psychology, 99(3), 445 to 476.

Kalyuga, S., Ayres, P., Chandler, P. and Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23 to 31.

Kalyuga, S., Chandler, P. and Sweller, J. (2001). Learner experience and efficiency of instructional guidance. Educational Psychology, 21(1), 5 to 23.

Kalyuga, S. and Singh, A.-M. (2016). Rethinking the boundaries of cognitive load theory in complex learning. Educational Psychology Review, 28(4), 831 to 852.

Kapur, M. (2008). Productive failure. Cognition and Instruction, 26(3), 379 to 424.

Karpicke, J. D. and Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966 to 968.

Kirschner, P. A., Sweller, J. and Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75 to 86.

Likourezos, V. and Kalyuga, S. (2017). Instruction-first and problem-solving-first approaches: Alternative pathways to learning complex tasks. Instructional Science, 45(2), 195 to 219.

Miller-Cotto, D. and Medrano, J. (2026). Does working memory moderate the effect of fading on math performance?. British Journal of Educational Psychology, 96(1), 108 to 126.

Pearson, P. D. and Gallagher, M. C. (1983). The instruction of reading comprehension. Contemporary Educational Psychology, 8(3), 317 to 344.

Renkl, A. and Atkinson, R. K. (2003). Structuring the transition from example study to problem solving in cognitive skill acquisition: A cognitive load perspective. Educational Psychologist, 38(1), 15 to 22.

Rosenshine, B. (2012). Principles of instruction: Research-based strategies that all teachers should know. American Educator, 36(1), 12 to 19, 39.

Schwonke, R., Renkl, A., Krieg, C., Wittwer, J., Aleven, V. and Salden, R. (2009). The worked-example effect: Not an artefact of lousy control conditions. Computers in Human Behavior, 25(2), 258 to 266.

Sherrington, T. (2025). Exploring confusion around modelling: "I do, we do, you do" vs "Guided and Independent Practice". teacherhead, 21 November.

Stockard, J., Wood, T. W., Coughlin, C. and Rasplica Khoury, C. (2018). The effectiveness of Direct Instruction curricula: A meta-analysis of a half century of research. Review of Educational Research, 88(4), 479 to 507.

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257 to 285.

Sweller, J. and Cooper, G. A. (1985). The use of worked examples as a substitute for problem solving in learning algebra. Cognition and Instruction, 2(1), 59 to 89.

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Webb, S., Massey, D., Goggans, M. and Flajole, K. (2019). Thirty-five years of the gradual release of responsibility: Scaffolding toward complex and responsive teaching. The Reading Teacher, 73(1), 75 to 83.

Wineburg, S. S. (1991). Historical problem solving: A study of the cognitive processes used in the evaluation of documentary and pictorial evidence. Journal of Educational Psychology, 83(1), 73 to 87.

Wood, D., Bruner, J. S. and Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17(2), 89 to 100.

Wu, L., Valcke, M. and Van Keer, H. (2023). Differential effects of reading strategy intervention for three levels of comprehenders: Focus on text comprehension and autonomous reading motivation. Learning and Individual Differences, 104, 102290.