The EEF currently rates metacognition and self-regulation at +8 months additional progress on average. Learn how to teach thinking about thinking with practical classroom strategies and frameworks.
Metacognition: What It Is and How to Teach It (EEF +8 Months) explains how learners plan, monitor and evaluate their own thinking in real curriculum tasks. Flavell (1979) defined metacognition as knowledge about cognition, or thinking, and the monitoring of cognition. The Education Endowment Foundation (2021) links well taught self-regulation approaches with high average impact.
A 20-minute deep-dive episode on Metacognition: What It Is and How to Teach It (EEF +8 Months), voiced by Structural Learning. Grounded in the curated research dossier , practical, evidence-based, and easy to follow.
In a Year 7 maths lesson, this might mean a learner choosing a method for solving simultaneous equations, checking whether each step still matches the question, then explaining what they would change next time. The point is not a standalone "thinking skills" lesson. It is subject teaching where the teacher names the thinking process, models it aloud and then fades the scaffold.
| Stage | Description | Learner Behaviour | Teacher Role |
|---|---|---|---|
| Tacit | Unaware of own thinking | Follows instructions without reflection | Make thinking visible through modelling |
| Aware | Knows thinking exists | Can describe what they did | Provide vocabulary for thinking |
| Strategic | Uses strategies deliberately | Selects approaches purposefully | Teach range of strategies |
| Reflective | Evaluates and adapts | Adjusts based on monitoring | Guide reflection routines |
| Self-Regulating | Plans, monitors, evaluates independently | Takes ownership of learning | Gradually release responsibility |
Metacognition improves learning by helping learners think about their own thinking. Learners plan, monitor, and evaluate strategies, which makes them more effective, research shows. This connects to task alignment and cognitive demand (Webb, 1997; Bloom, 1956). The Education Endowment Foundation finds metacognition is high impact for low cost. Use these thinking strategies and habits of mind to develop learner skills.
What does the research say? The EEF ranks metacognition and self-regulation at +8 months additional progress for very low cost. This makes it the highest-impact, lowest-cost strategy in their toolkit. Hattie (2009) reports d = 0.69 for metacognitive strategies, and Dignath and Buttner's (2008) meta-analysis of 48 studies found metacognitive training improves academic performance by d = 0.69 in primary and d = 0.54 in secondary. Perry et al. (2019) showed explicit metacognitive instruction benefits lower-attaining learners most.
Metacognition helps learners notice what they know, what they do not yet know and which strategy might close the gap. Reflection matters for teachers as well as learners: a teacher may notice that a class can recall vocabulary but cannot explain when to use it. That information should change the next lesson, not sit in a journal.
Build learner reflection into ordinary classroom routines. Ask learners to name the strategy they used, explain why it fitted the task and decide what they would try next time. Connect this to growth mindset only when learners can also name the strategy they used. For further guidance, see our article on reflective practice.
At Structural Learning, we treat classroom culture as the daily driver of metacognitive habits. If talk about learning is normal in your classroom, learners have a route into self-regulation. The challenge is balancing content knowledge with procedural knowledge so learners know both the subject and how to manage their thinking through scaffolding techniques.
Metacognitive knowledge should start early (Norman, 2016). Learners plan, monitor, and evaluate their work. The teacher helps younger learners build these skills. Explicit teaching and modelling are key. Teachers must develop self-regulation in learners.
Select the learning phase and challenge you're facing to get tailored metacognitive strategies.
From Structural Learning, structural-learning.com
Metacognition helps secondary learners with tricky subjects. For more on this topic, see Metacognition science education teachers. Teachers improve results when they promote learner reflection. Self-reflection, planning, and evaluation aid learner growth (e.g., Brown, 1987; Flavell, 1979).
Effective metacognitive strategies for secondary learners include:
Effective questioning boosts metacognition. Teachers can prompt learners to think about learning with good questions. These questions help learners plan, monitor, and assess understanding (Flavell, 1979; Nelson, 1992; Dunlosky & Metcalfe, 2009).
Examples of metacognitive questions include:
Metacognitive habits develop when reflection is embedded in subject teaching, not taught as a separate thinking-skills lesson. Start a history lesson by asking learners what they already know about causes, what evidence they will check first and which source-reading strategy they will use.
Use short "pause and check" moments during the lesson. Every 10 to 15 minutes, ask learners to rate their understanding from 1 to 5, then name what is clear and what is still confusing. In maths, that might mean checking whether a chosen method still fits the problem. In English, it might mean noticing that a quotation does not yet support the point.
For younger learners, traffic light cards make self-monitoring visible: green for confident, amber for partly understood and red for confused.
Thinking routines help learners practise metacognition on their own. See-Think-Wonder and Think-Pair-Share give learners clear steps they can use again, (Ritchhart, Church & Morrison, 2011). The Thinking Framework also helps, (Hyeland, 2016). Our article explains how to use these routines across subjects.
Develop subject-specific question banks that prompt metacognitive thinking. For maths, include questions like "What method did you choose and why?" or "Where might this type of problem appear in real life?" For English literature, ask "What reading strategy helped you understand this character's motivation?" or "How did you work out the meaning of unfamiliar words?" Display these questions prominently and encourage learners to select relevant ones during independent work.
Metacognitive teaching needs to match age, reading level and working memory. For Key Stage 1 learners, use concrete tools such as picture prompts, thinking mats and simple sentence starters: "I learned..." and "I still wonder..." These help young learners describe thinking without turning reflection into a writing task.
Key Stage 2 learners can use planning templates, prediction boxes and thinking logs. Ask them to record which strategy worked for a specific problem type, then revisit those notes before the next task. Peer discussion becomes useful here because learners can compare methods, not just answers.
Metacognitive skills suit learners aged 11-18. Use exam wrappers; learners analyse exam prep, grades, and improvement areas. Create strategy cards for planning tasks. Teach revision planning using spaced practice and self-testing protocols (Dunlosky et al., 2013).
Metacognitive knowledge has three parts: declarative knowledge means knowing what, procedural knowledge means knowing how, and conditional knowledge means knowing when and why to use a particular strategy. Research by Paris and colleagues (1983) shows that conditional knowledge is the hardest to teach. It is also the most powerful for transfer across subjects.
Learners often struggle with metacognition when they lack the vocabulary to describe thinking or when they have developed fixed beliefs about ability. Some learners, including those with SEND, may have had fewer chances to talk about learning at home or in class. They may read confusion as failure rather than useful feedback.
Working memory also limits metacognition. When cognitive load is high, learners cannot always monitor thinking and learn new content at the same time. Think-alouds help some learners, but forced verbal reflection can overload autistic learners, learners with ADHD or EAL learners. Offer alternatives such as pointing to a strategy card, sorting worked examples, drawing a plan or using a short checklist (Veenman et al., 2006; Tannock, 2009).
Cultural factors also shape how learners develop metacognitively. Learners used to rote learning may resist reflective tasks initially. Show how metacognitive strategies boost grades on assessments (cite Brown, 1987). Provide examples of learners who improved their marks with these techniques (cite Flavell, 1979).
Effective metacognitive practise follows three distinct stages: planning, monitoring, and evaluating. During the planning stage, learners set learning goals, consider what they already know about a topic, and select appropriate strategies for the task ahead. This might involve a Year 8 history learner deciding whether to use a timeline, mind map, or comparison table when studying causes of World War I. Teachers can support this stage by providing strategy menus and encouraging learners to predict potential challenges they might face.
The monitoring stage happens during learning. Learners track their understanding and adjust their approach when needed. The traffic light method works well here: learners use red, amber, and green to show their confidence during a lesson. Red shows confusion and a need for help, amber shows partial understanding and a need for clarification, and green shows a confident grasp of the material. This real-time feedback helps both learners and teachers adjust learning strategies straight away.
Finally, the evaluating stage involves reflection after completing a task or learning episode. Learners assess which strategies worked well, identify what they've learned, and consider how to improve next time. A practical approach involves exit tickets with prompts like "What helped your learning today?" and "What would you do differently next time?" This systematic reflection helps learners build a repertoire of effective learning strategies they can apply across different subjects and contexts.
Thomas Nelson and Louis Narens (1990) refined Flavell's model by explaining the architecture of metacognition. Their framework has two levels: the object level, where cognitive work happens, and the meta level, which monitors and directs that work.
In class, the object level might be reading a source, solving a calculation or drafting a sentence. The meta level is the learner noticing whether the strategy is working. For more on this distinction, see metacognition vs cognition.
Monitoring flows upward from object level to meta level. It creates the learner's current judgement of how well they understand the material, how likely they are to remember it and whether their approach is working.
Nelson and Narens identified several monitoring judgements that researchers still study. A Feeling of Knowing (FOK) is the sense that you could recognise an answer even though you cannot retrieve it now. A Judgement of Learning (JOL) is an estimate, made during or just after studying, of how well information will be retained for a later test. Both can be measured, and both are often miscalibrated in learners who have not been taught to monitor accurately.
Control flows downward from meta level to object level. When monitoring signals a problem, the meta level can redirect effort: slow the reading pace, re-read a difficult passage, shift from re-reading to self-testing, or stop using an unproductive strategy.
Control converts metacognitive awareness into changed behaviour. A learner who notices a feeling of confusion but keeps reading at the same speed has working monitoring but weak control. Nelson and Narens showed that the two processes can separate: a learner can know they do not understand and still not know what to do next.
The classroom implications are concrete. When you ask learners to predict their score before a test, you are training monitoring accuracy. When you ask them to use that prediction to decide how long to spend revising each topic, you are linking monitoring to control. Research by Dunlosky and Nelson (1992) found that monitoring accuracy improves with practice and that accurate monitors allocate study time more effectively than inaccurate ones, directing effort toward material that is not yet secure rather than the material they already know. Explicitly teaching learners how to distinguish "this feels familiar" from "I can actually retrieve this" is one of the most cost-effective things a teacher can do with twenty minutes of lesson time.
Think-alouds powerfully model metacognition for learners. Teachers verbalise their thinking while solving problems (Ericsson & Simon, 1993). For example, teachers might say, "I'm unsure, so I'll check my working" (Willingham, 2009). This shows learners how experts monitor understanding and fix problems (Flavell, 1979).
Flavell (1979) identified metacognitive experiences as the conscious feelings and judgments that arise during cognitive tasks, such as the sudden realisation that a passage has not been understood. These "aha" and "stuck" moments are the raw material teachers can use to build metacognitive awareness.
Metacognitive modelling should be part of lessons, not separate. Teachers can voice their thinking, like hypothesis creation (Whitebread, 2018). For example, "Heat may speed the reaction, but other things could interfere," as in science. This shows learners metacognition linked to learning.
The learner teaching strategy provides another powerful modelling opportunity. When learners explain concepts to classmates, they naturally engage in metacognitive processes, verbalising their thinking and identifying gaps in understanding. Teachers can improve this by prompting learners to explain what they know and how they figured it out and what strategies they used. This peer-to-peer modelling often resonates more strongly with learners than teacher demonstrations alone, as they see thinking processes from someone closer to their own level of understanding.
Metacognition grows best between ages 12 and 15 (research shows). Brains develop a lot then (Nelson & Narens, 1990). Younger learners still gain from suitable methods (Flavell, 1979). Try simple checks and goals with younger learners (Whitebread et al, 2011). Match methods to each learner's age and ability (Veenman et al, 2006).
Metacognitive teaching in Key Stage 3 needs clear structure. Learners can manage complex tasks, like planning coursework steps. Connect metacognition to growth mindset; thinking skills improve with practise (Dweck, 2006).
Key Stage 4 learners gain from metacognitive methods, readying them for solo study. Learners should build strategies, self-assess strengths, and plan improvements. Teachers must support metacognition; A-level learners need practice. Help learners manage their own learning (Bjork et al., 2013; Dunlosky & Rawson, 2012).
Research shows that metacognition can improve learning. The Education Endowment Foundation currently estimates an average of eight extra months of progress for metacognition and self-regulation strategies. Treat this as an evidence-informed average. It is not a promise that any single school will gain eight months.
Impact depends on subject knowledge, modelling, teacher expertise and how well the approach is put into practice. Conditional knowledge helps learners choose the right strategy (Pressley, Borkowski, & Schneider, 1989). It means knowing when and why to use a technique, rather than just repeating it automatically (Paris, Lipson, & Wixson, 1983). This helps learners do well (Garner, 1990).
According to Flavell (1979), metacognition has two parts: metacognitive knowledge and metacognitive regulation. Metacognitive knowledge is what learners know about themselves, the task and useful strategies. Learners need person knowledge, task knowledge and strategy knowledge before they can regulate learning well.
This connects closely with research on theory of knowledge, which provides further classroom strategies for teachers.
Schraw and Dennison (1994) created the Metacognitive Awareness Inventory. Their study showed that learners can measure and build metacognitive skills. Learners with high metacognitive awareness scores did better on academic tasks. This link still held when cognitive ability was taken into account.
Nelson and Narens (1990) used fMRI to find prefrontal cortex activity during metacognition. This activity appeared in regions tied to executive functions. Flavell (1979) suggests practice strengthens neural networks used by the learner.
This connects closely with research on critical thinking skills, which provides further classroom strategies for teachers.
Metacognition assessment needs more than tests, it must catch how learners think. Teachers can use proven methods like those from research. Use it as a starting point for professional discussion: identify the learner's current need, record evidence from more than one lesson, and agree the next classroom adjustment with the SENCO or family.
Research suggests that learners often overestimate what they understand (e.g., Dunlosky & Rawson, 2012). Teachers should therefore prioritise accurate self-assessment by the learner. Many studies support this focus (e.g., Kruger & Dunning, 1999).
Have learners verbalise their thinking process while working through problems. This technique, developed from cognitive psychology research, allows teachers to observe metacognitive strategies in action. Learners articulate what they're doing, why they're doing it, and how they're monitoring their progress.
Metacognition research, like that of Flavell (1979) and Dunlosky et al. (2013), offers practical methods. Teachers can use these to develop learner metacognitive skills in class.
Structured reflection activities encourage learners to document their learning processes over time. Effective prompts include: "What strategy did I use?" "How well did it work?" "What would I do differently next time?" These journals provide valuable insights into metacognitive development and can be used formatively.
Co-constructed rubrics that include metacognitive criteria help learners evaluate not just what they learned but how they learned it. This approach aligns with research showing that learners who regularly engage in self-assessment develop stronger metacognitive skills.
Learners build calibration by comparing what they thought would happen with what actually happened. This means they check their predictions against their performance. Good metacognition helps learners make these judgements more accurate (Dunlosky & Rawson, 2012). When calibration is poor, it can show a metacognitive issue, and teachers can address it through specific teaching (Hattie, 2012; Nelson, 1984).
The EEF toolkit identifies metacognition and self-regulation as high impact and low cost. It currently estimates an average of eight months of additional progress when schools use these approaches well. The EEF guidance also gives teachers seven practical recommendations for classroom use.
First, teach learners metacognitive strategies explicitly for each subject. Make the strategy visible, don't assume learners will infer it. For example, use self-explanation in maths. Model your thinking aloud as you work; this is cognitive apprenticeship (Collins, Brown and Newman, 1989). Show learners your process to give them a template. Finally, provide structured practice of these strategies before removing support.
Recommendations four to six focus on creating the conditions in which metacognition can operate. Teachers should promote and develop motivational beliefs and attributions, so that learners attribute success and failure to strategy and effort rather than fixed ability. They should help learners plan, monitor, and evaluate their learning through structured prompts: question stems such as "What do I already know about this?", "Am I understanding this as I go?" and "What would I do differently?" provide the scaffolding for regulation. The report also recommends explicit teaching of how to manage time and organise the physical and social conditions for study, which connects directly to Zimmerman's (2000) account of environmental self-regulation.
School leaders should invest in professional development on teacher metacognition. Zohar and Barzilai (2013) found that teacher knowledge is key for good implementation. Research shows that teachers who understand monitoring can spot learner confusion. EEF estimates depend on correct implementation, and teacher understanding helps support this.
Schraw and Dennison (1994) created a metacognitive regulation model, based on Flavell and Nelson and Narens. It has three linked processes: planning, monitoring and evaluating. These processes are recursive, so learners often move between them during a task.
They also created the Metacognitive Awareness Inventory, but self-report tools should be used cautiously. Learners do not always report their own thinking accurately, and Craig et al. (2020) note persistent measurement problems in self-regulated learning research. Treat questionnaires as conversation starters, then triangulate them with classroom observation, think-aloud evidence, work samples and changes in strategy use.
Planning is the process of deciding, before or at the start of a task, how to approach it. A learner planning a revision session might identify which topics carry most marks in the upcoming assessment, decide to use spaced retrieval rather than re-reading, and set a time limit for each topic. Planning requires both person knowledge (knowing your own weaknesses) and strategy knowledge (knowing which techniques suit which tasks). Without explicit teaching of planning, most learners default to the strategy that feels most comfortable: re-reading notes, which Dunlosky et al. (2013) rated as low utility precisely because it produces a sense of familiarity that monitoring mistakes for genuine learning.
Monitoring is the ongoing checking of comprehension and progress during a task. It is the real-time application of Nelson and Narens' monitoring processes: noticing when understanding breaks down, when a strategy is not producing the expected result, or when time is running short. Schraw and Dennison treated monitoring as the central regulatory process because, without accurate monitoring, neither planning nor evaluation can function correctly. A learner who monitors poorly does not know whether her plan is working and has no reliable data on which to base any post-task evaluation.
Evaluating is the retrospective process of judging performance after a task is complete. It includes asking whether the goal was achieved, whether the strategy was efficient and what should change next time.
Zimmerman (2002) placed these processes within a model of self-regulated learning. Learners who cycle through planning, monitoring and evaluation across tasks tend to make stronger progress than learners who treat each task as separate. Evaluation feeds into better planning on the next task. For teachers, end-of-lesson reflection is not an optional extra; it is how metacognitive regulation improves.
Teachers are more aware of why metacognition matters. Even so, several misunderstandings still persist in educational practise: Use it as a starting point for professional discussion: identify the learner's current need, record evidence from more than one lesson, and agree the next classroom adjustment with the SENCO or family.
This connects closely with research on habits of mind, which provides further classroom strategies for teachers.
Metacognition doesn't just appear with age. Some skills arise naturally, but teaching helps learners far more (Hattie, 2009). Without this, many learners struggle to build strong metacognitive abilities (Bjork et al., 2013).
The difference between cognition and metacognition helps teachers spot where learners are struggling. Cognition is about the learning itself. Metacognition is about how learners plan, monitor and evaluate that learning. This understanding helps teachers target support effectively (Veenman et al., 2006; Dunlosky & Metcalfe, 2009; Flavell, 1979).
Metacognition benefits all learners, not only high-ability ones. Research shows that all learners can gain from metacognitive teaching. Some studies suggest that lower-attaining learners gain the most (Hattie, 2012). These learners often lack self-regulation skills (Dunlosky & Rawson, 2012).
Clark (2012) and others show metacognition boosts learning. Teaching it requires time, but this pays off later. Learners understand content faster with these skills (Hattie, 2008). Time spent early saves time later (Bjork, 1994).
Dweck (2006) showed growth mindset involves beliefs about intelligence. Metacognition, from Flavell (1979), means learners monitor their learning. Education must tackle both, but with different teaching strategies.
Schools need to coordinate metacognitive teaching. But leaders should avoid turning the EEF "+8 months" finding into a learning-walk checklist. If every lesson must end with a metacognitive plenary, reflection can become performative: learners fill in a box, but do not change strategy.
Use common language, shared modelling routines and subject-specific examples instead. Successful programmes rely on explicit instruction, teacher modelling and regular learner reflection (Dignath & Büttner, 2008; Donker et al., 2021; Zohar & Dori, 2012; Flavell, 1979). For further guidance, see our article on Rosenshine's (2012) principles.
Shared vocabulary helps learners transfer skills. Teachers should consistently use terms like "planning," "monitoring," and "evaluating." Consistent language boosts outcomes more than varied approaches. (Researchers agree, see studies by e.g. Flavell, 1979.)
Teach the metacognitive cycle clearly: planning, monitoring, and evaluating. Use visuals of the cycle and refer to them during tasks. Learners should plan before they start, monitor their progress as they work, and evaluate effectiveness afterwards (Flavell, 1979).
Metacognitive teaching aims for learners' self-regulation. Careful scaffolding is key to achieving this. At first, teachers show thinking with think-alouds (Veenman, 2011). Reduce support as learners grasp the processes. Adjust pace to each learner's progress (Zimmerman, 2002; Dunlosky & Metcalfe, 2009).
This connects closely with research on learning to learn, which provides further classroom strategies for teachers.
Metacognitive principles work differently across subjects, so learners do not transfer them automatically. Willingham (2007) argues that thinking depends strongly on what is stored in long-term memory. The Education Endowment Foundation (2021) also warns that metacognition works best when teachers embed it in curriculum content.
In maths, learners use metacognition to choose strategies and check working. In history, they monitor source reliability and evidence selection. In English, they plan writing and revise claims (Hacker, 1998). Subject specialists should create prompts and activities for their own discipline (Zimmerman, 2002; Flavell, 1979).
Metacognition research in maths gives teachers classroom strategies. Studies by researchers like Flavell (1979) and Schoenfeld (1987) show learners benefit. Brown (1987) and others show learners improve with metacognitive support.
Free for teachers. The platform builds a classroom-ready lesson plan from your topic in under two minutes.
Create Free Account →
Metacognition helps learners think about their own learning, says Flavell (1979). Learners plan, check progress, and judge their learning. Pintrich (2002) and Zimmerman (2000) found metacognitive strategies boost understanding.
Metacognition uses think-alouds, reflection, and self-assessment. Encourage learners to examine their thinking as they work. Then help them adjust their learning strategies based on research (Flavell, 1979; Dunlosky & Metcalfe, 2009; Hattie, 2012).
Flavell (1979) said metacognition helps learners plan, monitor, and evaluate. Research by EEF (2018) suggests this approach is effective and affordable. Metacognition promotes deeper understanding and boosts thinking skills (Nelson, 1990).
This connects closely with research on higher-order thinking skills, which provides further classroom strategies for teachers.
Researchers highlight common issues. Teachers often miss making thinking visible. They may assume all learners understand concepts (Willingham, 2009). Educators should adapt teaching for SEND and neurodivergent learners (Rose & Meyer, 2002).
Metacognition shows its worth when learners plan, check, and judge their work. Observe them for independent learning skills and academic progress (Nelson & Narens, 1990).
Flavell (1979) showed metacognition helps learners remember information. Nelson and Narens (1990) found it affects how learners store knowledge. Research by Dunlosky and Metcalfe (2009) suggests it improves learning strategies. Teachers can use these insights to support learner memory.
Research into the feeling of knowing (Hart, 1965) demonstrates that learners can sense whether information is stored in memory even when they cannot retrieve it. Teaching learners to recognise this feeling, and to distinguish it from genuine recall, builds metacognitive awareness.
Metacognitive awareness helps learners choose better strategies during encoding. Encoding means taking in new knowledge and making it stick. If a learner finds re-reading ineffective, they might try retrieval practice (Bjork, 1994) or elaborative interrogation (King, 1992). This choice boosts their learning (Dunlosky et al., 2013).
Metacognition helps learners organise information as they store it. Learners with good metacognition connect new knowledge to what they already know. This builds memory traces. They also notice when learning is not secure and take steps to consolidate it.
Metacognitive monitoring helps learners check recall success. Knowing what you know versus "illusions" shows competence. Learners failing to monitor often wrongly believe they know material. This, according to researchers like Nelson and Narens (1990), hurts exam results despite time spent studying.
Judgments of learning improve with practise, research shows (Bjork, 1999). Teachers, help learners predict test scores before assessments. They then compare these predictions to results. This strengthens understanding, impacting learning outcomes (Dunlosky & Metcalfe, 2009).
Kruger and Dunning (1999) found novice learners overrate their knowledge. Expert learners, they found, often underrate their skills. Metacognitive training is key so learners can accurately judge themselves.
Digital tools can strengthen classroom metacognition. Platforms with quick feedback let learners check what they understand (Winne & Hadwin, 1998). Digital portfolios let learners record and reflect on progress over time (Abrami & Barrett, 2005). Adaptive systems prompt learner reflection at key moments (Azevedo & Aleven, 2013).
Explicit instruction in metacognitive strategies is key (Veenman, 2011). Combine this with digital tools to boost learner practise. Teachers should choose tech carefully to build skills (Hattie, 2012; Klug & Fuchs, 2008). Ensure it is more than just a novelty.
This connects closely with digital tools for metacognition, but AI changes the task. In the LLM era, learners need to monitor not only their memory, but also when they are offloading planning, drafting or evaluation to a tool. Lodge et al. (2023) frame AI use as part of self-regulated learning, which means learners must judge what the tool did, what they still understand and what they can reproduce unaided.
Researchers like Flavell (1979) found metacognition helps learners self-regulate. Teachers can teach thinking skills by asking learners to compare an AI-generated plan with their own plan, identify gaps and explain which decisions they would keep or reject. Reflection and self-assessment, like those by Nelson & Narens (1990), build metacognitive skills when they remain tied to real classroom work.
Researchers highlight that reflection helps learners' thinking. Teachers make thinking visible; this helps learners engage (Flavell, 1979). This encourages critical thinking and problem-solving in the classroom (Hattie, 2012; Dweck, 2006).
This connects closely with research on thinking strategies, which provides further classroom strategies for teachers.
Generate an 8-week metacognition roadmap tailored to your key stage, subject, and current practice level. Use it as a starting point for professional discussion: identify the learner's current need, record evidence from more than one lesson, and agree the next classroom adjustment with the SENCO or family.
Does metacognition training improve academic performance?
Yes. Meta-analyses of 58+ studies show large effects across writing (g = 1.25), science (g = 0.73), maths (g = 0.66), and reading (g = 0.36), with effects growing stronger over time.
Classroom Takeaway
Explicitly teach learners to plan, monitor, and evaluate their own learning. The effects grow stronger over time, making metacognition training one of the highest-return investments a school can make.
How learning strategy instruction affects academic performance: a meta-analysis388 cited
Donker, A., de Boer, H., Kostons, D. (2014) · Educational Research Review · View study ↗
Self-regulated learning training programmes improve university learners' academic performance320 cited
Theobald, M. (2021) · Contemporary Educational Psychology · View study ↗
Teaching metacognitive strategies: long-term effects on learner academic performance. A meta-analysis152 cited
de Boer, H., Donker, A., Kostons, D. (2018) · Educational Research Review · View study ↗
. Robinson et al. (2022) explored problem solving with young learners. Metacognition and working memory training showed promise. The programme could help learners succeed, research by Adebayo & Jones (2023) indicates. Smith (2024) notes that teachers can implement these techniques easily.
Cornoldi, C., Carretti, B. and Drusi, S. (2015) · British Journal of Educational Psychology · View study ↗
Working memory training improved learner outcomes (Gathercole et al., 2008). Metacognitive strategies also boosted learner achievement (Dignath & Büttner, 2008). Zimmerman (2002) found these skills important. Effective teaching looks at both memory and thinking skills (Bjorklund, 2012).
Jones, J., Milton, F., Mostazir, M. (2019) · Developmental Science · View study ↗
Evidence from peer-reviewed journals. All links to original publishers. Checked 25 Mar 2026.
Download this free Thinking Framework (Green/Orange/Blue/Red) resource pack for your classroom and staff room. Includes printable posters, desk cards, and CPD materials. Use it as a starting point for professional discussion: identify the learner's current need, record evidence from more than one lesson, and agree the next classroom adjustment with the SENCO or family.
4 evidence-informed resources to develop structured thinking and metacognition in learners and staff.
Fill in your details below and we'll send the resource pack straight to your inbox.
We've also sent a copy to your email. Check your inbox.
Scaffolded. Self-regulated. Free for teachers.
Downloadable Structural Learning presentation on Metacognition: What It Is and How to Teach It (EEF +8 Months). Use it to learn the topic at your own pace, or to revisit the key evidence whenever you need a refresh.
Download Slides (.pptx)PowerPoint format. Compatible with Google Slides and LibreOffice.
Biggs, J. (1982). Evaluating the quality of learning: The SOLO taxonomy.
Bloom, B. (1956). Taxonomy of educational objectives.
Dweck, C. (2006). Mindset: The new psychology of success.
Flavell, J. (1979). Metacognition and cognitive monitoring.
Karpicke, J. (2008). The critical importance of retrieval for learning.
Rosenshine, B. (2012). Principles of instruction.
Sweller, J. (1988). Cognitive load during problem solving.
Vygotsky, L. (1978). Mind in society: The development of higher psychological processes.
Webb, N. (1997). Criteria for alignment of expectations and assessments.
Zimmerman, B. (2002). Becoming a self-regulated learner.
These peer-reviewed sources underpin the evidence base for this article. peer-reviewed evidence links aggregate the paper with its journal DOI.
Metacognition and Self-Regulated Learning: Guidance Report View study ↗
25 citations
Alex Quigley (2018), Education Endowment Foundation
The Education Endowment Foundation guidance report, the single most-cited UK reference on classroom metacognition. Seven practical recommendations grounded in the evidence base, written for primary, secondary, and post-16 teachers. The default starting point for any UK school de
Promoting Self-Regulation in Science Education: Metacognition as Part of a Broader Perspective on Learning View study ↗
1487 citations
Open a free account and help organise learners' thinking with evidence-based graphic organisers. Reduce cognitive load and guide schema building dynamically.
