Metacognition in the Classroom

Metacognition helps learners plan, monitor and evaluate their learning with more care. The Education Endowment Foundation currently estimates an average eight months of additional progress for metacognition and self-regulation approaches. Flavell (1979) framed metacognition as knowledge and regulation of thinking. In classrooms, the challenge is to make that process clear without turning it into vague advice.

This guide links the evidence to concrete classroom routines.

Metacognition means the knowledge and control learners have over their own thinking. It includes how they plan a task, check their understanding while they work, and judge how well their strategy worked afterwards (Flavell, 1979).

‍

What Is Metacognition?

Metacognition means thinking about thinking. The term was introduced by John Flavell (1979), a developmental psychologist at Stanford, who defined it as "one's knowledge concerning one's own cognitive processes or anything related to them." In plain terms, it is the mental activity we use to oversee, direct, and evaluate our own learning. A learner who realises they have not understood a paragraph and decides to reread it is using metacognition. A learner who copies notes without noticing they cannot explain them is not.

Flavell (1979) noted three parts of metacognition. Metacognitive knowledge is what a learner knows about learning strategies. It includes task types and what works best (Flavell, 1979).

Metacognitive regulation means actively managing thinking during work, such as planning and checking progress (Flavell, 1979). Metacognitive experience is a learner's real-time feeling about how learning progresses (Flavell, 1979).

Metacognitive knowledge, skills and experience work together (Flavell, 1979). If metacognitive knowledge is weak, learners may choose poor strategies (Bjork (Bjork, 1994) et al., 2013). Even with good knowledge, poor monitoring can mean learners lose focus (Nelson & Narens, 1990).

So, teach all parts of metacognition (Zimmerman, 2000). Do not rely only on end-of-lesson reflections.

‍

Why Metacognition Matters for Learning

The EEF toolkit identifies metacognition and self-regulation as a high-impact, low-cost strand. It currently estimates an average eight months of additional progress. The main classroom message is not a single trick. Teachers should explicitly teach planning, monitoring and evaluation across subjects.

Hattie (2009) found self-reported grades had a very high effect size (d = 1.44). This shows learners' awareness links to achievement. Learners knowing their strengths study better. They also persevere and use feedback well.

Dunlosky et al. (2013) found that elaborative interrogation, self-explanation, and practice testing help learners learn. Dunlosky et al. (2013) also showed that rereading and highlighting often lead to poor recall. Even so, learners often choose weaker revision methods.

The cost of poor metacognition is most visible in assessments. Learners who revise by reading through notes often feel confident before an exam. They then perform worse than they expected. Learners who revise using retrieval practice feel less confident during revision but do better in the exam.

Koriat and Bjork (2006) called this gap between felt learning and real learning an illusion of knowing. Teaching metacognition means teaching learners not to trust the feeling of fluency as proof that they have learned.

‍

The Three Components of Metacognition

Flavell (1979) identified three types of metacognitive knowledge. Person knowledge is what a learner thinks about their learning strengths. Task knowledge involves understanding different demands of varied tasks. Strategy knowledge means knowing which approaches work best (Flavell, 1979).

A concrete classroom example: in a Year 10 history lesson on the causes of the First World War, a learner with strong person knowledge recognises that they understand the political causes but are shakier on the economic ones. Strong task knowledge tells them that the essay format will require them to weigh and compare causes, not just list them. Strong strategy knowledge leads them to use a comparison graphic organiser to clarify the relationship between causes before drafting. Without any one of these three, the learner's preparation is less efficient.

Learners use metacognition during lessons. Planning means using prior knowledge and choosing strategies. Monitoring means checking understanding and noticing when a strategy is not working.

Evaluating means judging the quality of work and thinking back on the strategy choice. The Education Endowment Foundation guidance report uses this plan-monitor-evaluate cycle as a practical classroom frame.

Focus more on metacognitive experience in class. Flavell (1979) said this includes feelings about learning. Efklides (2006) showed that confusion can mean a learner's method is not working.

Teach learners to see confusion as helpful data. Dunlosky & Metcalfe (2009) suggested modelling problem-solving, so learners can see how to respond when they get stuck.

‍

Teaching Metacognitive Strategies

The Education Endowment Foundation guidance report gives seven recommendations for teaching metacognition and self-regulated learning. Its practical message is clear. Teachers should model thinking, set the right level of challenge, support metacognitive talk and teach transfer directly. Metacognition should not be a loose request to reflect.

Planning strategies are often the most neglected of the three phases. Before a task, teach learners to ask: what do I already know about this? What do I not know? What does this task actually require me to do?

They can also ask: what approach will I use, and why? A simple pre-task planning template can support this. In a science lesson, a teacher can spend two minutes before a written explanation task asking learners to write one thing they know confidently, one thing they are unsure of, and one strategy they will use to check their explanation is accurate. This activates metacognitive knowledge before learners begin, rather than leaving them to start without reflection.

Learners often miss checks on understanding, which stops them from monitoring their learning. Willingham (2009) says instructions need comprehension checks. Rosenshine (2012) advises teachers to pause and ask learners what confuses them.

Red, amber, green cards can show learner confidence. Naming confusion helps learners build metacognitive skills.

Evaluation at the end of tasks needs to move beyond "did I finish?" to "did I understand, and how do I know?" A useful post-task prompt sequence is: what was the hardest part of this task? What strategy did I use, and did it work? What would I do differently next time?

These three questions link directly to metacognitive evaluation. For written work, learners can annotate their own drafts before teacher feedback. They can identify sentences they are unsure about. This helps them compare their self-assessment with the teacher's feedback and build metacognitive accuracy over time.

Kruger and Dunning (1999) showed that new learners often overrate their knowledge. Expert learners often underrate their abilities. This means good self-assessment relies on metacognition.

‍

Metacognition Across the Curriculum

Bjork (1994) found that metacognition strategy transfer can be weak. Learners may struggle to plan, monitor, and evaluate when they move between subjects. Nelson (1996) suggested that integration helps learning. Dunlosky (2009) recommends using relevant language and examples.

Learners check their understanding of maths problems before calculating (Schoenfeld, 1992). They can draw diagrams or note key information (Pape, 2004). "Worked example comparison" tasks help learners (Star et al., 2015). Learners compare solutions to find errors, which aids evaluation.

In English, metacognition fits well with reading comprehension and writing planning. Before learners read a challenging text, teachers can model "reading like a detective": what do the title and subheadings tell me? What genre is this, and what does that mean for how I read it? During reading, the "say something" technique asks learners to pause at marked points and say what they understand so far, what surprised them, and what they think will come next. These are metacognitive monitoring behaviours made concrete and social.

Metacognition matters for science learners. They must know "explain why" questions need causal reasoning. Learners should grasp that recall tasks differ from application tasks (Zohar, 1999). Thinking strategies let learners plan and monitor their learning more precisely (Whitebread et al., 2015; Krätzig & Arbuthnott, 2006).

Weinstein et al. (2000) found metacognitive planning helps learners choose a stance before reading. Learners check sources for support, problems, or contradictions. Perry (1998) and Atherton (2003) showed annotations clarify this, aiding teaching and argument building.

‍

Metacognition and Self-Regulated Learning

Zimmerman (2002) describes self-regulated learning as forethought, performance, and self-reflection. These steps are similar to planning, monitoring, and evaluation. Zimmerman (2002) also noted that motivation drives self-regulation. Self-efficacy, goals, and outcome reactions all influence this process.

Learners need both metacognition and motivation to succeed. The current Education Endowment Foundation toolkit reports high average impact for metacognition and self-regulation. It also warns that this impact depends on explicit teaching and good implementation. Keep growth mindset separate: it can be useful classroom language, but it should not be presented as the EEF's core route to metacognitive improvement.

Dweck (2006) and Flavell (1979) provide teachers with classroom ideas. Growth mindset and metacognition boost learner progress.

Brown (1987) showed that executive function, including working memory, helps learners manage their own learning. When memory is weak, learners may lose track of goals. Clear teaching structures reduce cognitive load, or the mental effort of a task. Checklists also help learners use successful strategies (Bjork, 1994; Flavell, 1979).

This connects closely with research on theory of knowledge, which provides further classroom strategies for teachers.

Schraw and Dennison (1994) made the Metacognitive Awareness Inventory. It checks older learners' metacognitive knowledge and control. Their work showed that these skills can grow at different speeds.

For example, a learner may know retrieval helps but reread anyway. Teaching should target both areas to support better learning.

Hart's (1965) research shows that learners can sense stored information, even when they cannot recall it. Teaching learners to recognise this feeling builds metacognitive awareness. It also helps them tell the difference between a feeling of knowing and actual recall.

‍

Thinking Tools That Develop Metacognition

Structured tools are the most reliable way to make metacognitive processes visible and teachable. When thinking is invisible, neither the teacher nor the learner can see where it is going wrong. When it is made explicit through a graphic organiser, a thinking routine, or a self-assessment prompt, both can observe, discuss, and improve it.

Graphic organisers are the most flexible metacognitive tool. A KWL chart (Know, Want to know, Learned) helps learners plan a task and then judge what they have learned. A Cornell notes template, with a summary box at the bottom, asks learners to shorten and review their own notes.

A comparison table asks learners to set out their own criteria before they fill it in. These tools have metacognitive value because learners must show their thinking before, during, or after a task. This creates a record that they can review and improve.

Project Zero routines improve learner thinking, research suggests. "See-Think-Wonder" helps learners observe carefully before forming ideas (Harvard University). "Think-Puzzle-Explore" uses knowledge to make learners question things (Project Zero). "I Used to Think, Now I Think" allows learners to consider changes in understanding. These routines give clear thinking steps, so you don't need to invent your own (Project Zero).

Thinking maps add a further layer: each map type corresponds to a specific cognitive task. A circle map (brainstorming context), a tree map (classifying), a brace map (analysing part-whole relationships). Teaching learners which map to choose for which task is itself metacognitive strategy knowledge. A learner who reaches for a flow map when they need to sequence a process, rather than drawing random boxes and arrows, is applying metacognitive knowledge about task demands and appropriate tools.

Metacognitive knowledge includes 'what', 'how', and 'when/why' (Paris et al., 1983). Learners find conditional knowledge (when and why) difficult. However, Paris et al. (1983) found it best supports knowledge transfer.

Structured reflection journals, used weekly rather than as a one-off task, build metacognitive evaluation as a habit. The most effective prompts are specific: "Write one thing you understood in today's lesson that you would be able to explain to someone else without looking at your notes. Write one thing you are still unsure about. Write what you plan to do before next lesson to address the gap." Generic prompts ("What did you learn today?") produce surface responses; specific prompts produce metacognitive thinking.

This connects closely with research on habits of mind, which provides further classroom strategies for teachers.

‍

Assessing Metacognitive Growth

Metacognition is harder to assess than subject knowledge, but teachers can still observe it. The most reliable indicators are behavioural: does the learner pause before a complex task and seem to be planning? Do they check their work mid-task?

Do they ask specific questions about what the task requires, rather than just beginning? Do they revise their approach when an initial strategy is not working? These observable behaviours are stronger indicators of metacognitive regulation than self-report questionnaires alone.

Flavell (1979) showed that metacognition helps learners. Schraw and Dennison (1994) gave teachers useful ways to teach it. Together, this research gives teachers practical classroom strategies. It helps them begin metacognition activities with learners.

Calibration checks are a simple way to assess learners before and after a task. Before learners start, ask them to rate their confidence from 1-3 (Bjorkman, 1994). After the task, they compare that rating with their result.

If confidence and results do not match, this shows a calibration gap (Hacker et al, 2000). When learners judge their confidence more accurately over time, it shows metacognitive growth (Winne & Hadwin, 1998).

A judgment of learning (JOL) is a learner's prediction of how well they will remember material on a future test. Nelson and Narens (1990) found that delayed JOLs, made after a short gap rather than immediately, are far more accurate and help learners calibrate their revision effort.

Use think-alouds to check learner metacognition (Veenman, 2011). In a think-aloud, learners say their thoughts as they solve problems. Listen for words that show planning, monitoring, and evaluating.

If learners do not use this language, it can signal weak metacognitive skills (Zimmerman, 2000). Model think-alouds before learners try them (Flavell, 1979).

Portfolio assessment shows learner progress well, unlike tests. Planning sheets and self-assessments help learners monitor their own learning (EEF, 2018). Teachers note learners become more aware of their learning when they assess metacognition.

‍

Common Misconceptions About Metacognition

Flavell (1979) says metacognition actively manages thinking. It goes beyond simple reflection. For example, learners who list lesson points are showing reflection. 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.

Nelson and Narens (1990) think reflection alone may limit gains. Even so, end-of-lesson reflection can help learners consolidate knowledge.

A second misconception is that younger children cannot benefit from metacognitive instruction. The research does not support this. Brown et al. (1983) showed that children as young as five can learn to check their own understanding during story reading. EEF evidence also shows clear gains for primary-age learners when teachers teach metacognitive strategies.

This teaching needs to be concrete and rich in language. Teachers should model metacognitive language clearly. "I'm going to think out loud so you can hear what I'm doing in my head" is accessible to a Year 1 class. The ideas are simpler, but the principle is the same.

Flavell (1979) thought metacognition depends on the subject. Learners can find English techniques hard to use in maths. Explaining concepts changes between subjects. Subject-specific teaching works better than general skills (researchers).

This connects closely with research on learning to learn, which provides further classroom strategies for teachers.

Self-marking can teach learners to think about their learning. Sadler (1998) found unguided self-marking assesses superficially. White & Frederiksen (1998) say learners judge reasoning, not just right answers. Andrade & Valtierra (2001) showed structured methods boost thinking skills better.

Ennis (1993) and Halpern (1998) suggest practical strategies for critical thinking. Teachers can use these to help learners develop key reasoning skills. Willingham's (2007) research shows prior knowledge boosts learner thinking.

‍

Building a Metacognitive Classroom Culture

Without a supportive classroom culture, individual strategies show limited results. Learners need classrooms where admitting "I don't understand" sparks curiosity. Changing minds with new evidence should be intellectual strength. Frame difficulty as a normal part of learning, not a failure (Costa & Kallick, 2008; Dweck, 2006).

Classroom language is the most direct tool for building this culture. Teachers who consistently use metacognitive language model the behaviour they want learners to adopt. "I'm going to think about what I already know before we start" is metacognitive planning. "I noticed I made an error here and I want to understand why" is metacognitive evaluation. "This is confusing me, so I'm going to try a different approach" is metacognitive monitoring. When learners hear this language daily, they acquire the vocabulary and the habits that go with it. This is especially important for learners from disadvantaged backgrounds, who are less likely to have encountered this language at home.

Classroom routines that include metacognitive moments are more reliable than activities that rely on teacher memory. Use a two-minute planning prompt before every extended writing task. Add a mid-task check-in question to every practical activity. Before moving on, ask, "what was hard and what did you do about it?"

When these pauses are built into lesson design, they become normal and expected. They should not depend on whether there is time left. Learners who meet them often, across subjects and years, develop real habits. They are less likely to give performed responses.

Teachers should focus on scaffolding and metacognition. Scaffolding means giving support, but it should not do the work for learners. When support replaces learner effort, it creates dependency.

Teachers should prompt learners to monitor themselves, as in work by Vygotsky (1978). Scaffolding that supports effort builds independence, like in Wood et al.'s (1976) research. Good teaching gradually reduces support as learners gain skill.

Hattie (2009) linked metacognition with self-regulation. The Education Endowment Foundation guidance report can help departments use a shared language for planning, monitoring and evaluating learning. Teachers do not need a separate programme, but they do need steady routines, clear modelling and chances for learners to explain their strategy choices.

Thinking frameworks can improve learners' metacognition. De Bono's Six Thinking Hats (de Bono) helps learners look at evidence and risks. This can help them move beyond their usual thought patterns. Bloom (Bloom, 1956)'s Taxonomy, from recall to evaluation, offers similar benefits.

What makes the difference between teachers who see genuine metacognitive growth in their classes and those who do not is rarely the choice of specific strategy or tool. It is consistency, explicitness, and the willingness to value the process of thinking as much as its products. Learners learn to think about their thinking when teachers persistently, visibly, and without embarrassment think about theirs.

‍

‍

Limitations and Critiques

Metacognition is strongly supported, but it is not a generic skill that transfers neatly across subjects. Willingham (2007) and Tricot and Sweller (2014) argue that learners need enough domain knowledge before they can judge the quality of their own reasoning. A learner cannot monitor a history essay well if they lack the historical facts needed to test the argument.

There is also a measurement problem. Veenman (2011) warned that self-report surveys often miss the real-time monitoring learners use during tasks. Craig et al. (2020) raised similar concerns about relying on broad questionnaires to measure self-regulated learning. This matters because headline claims can rest on data that do not fully capture classroom behaviour.

Metacognitive routines can also create an executive function tax. Long reflection logs, dense planning sheets and abstract prompts may disadvantage ADHD, autistic and high-cognitive-load learners if they add demands without reducing task complexity. Cultural limits matter too: many studies reflect Western assumptions about independent learning, verbal reflection and individual goal setting.

Finally, generative AI can weaken monitoring if learners outsource planning, drafting and checking before they have struggled with the task themselves. Despite these limits, metacognition remains valuable when it is taught through subject knowledge, modelled clearly, and kept light enough to support learning rather than distract from it.

‍

References

Bjork, R. (1994). Memory and metamemory considerations.

Bloom, B. (1956). Taxonomy of educational objectives.

Flavell, J. (1979). Metacognition and cognitive monitoring.

‍

Further Reading: Key Research Papers on Metacognition

These five studies and reports form the core evidence base for metacognitive instruction in classroom settings. Each is directly relevant to practising teachers.

Flavell (1979) introduced metacognition as knowledge and regulation of thinking. Nelson and Narens (1990) explain cognitive monitoring, and Pintrich (2002) connects metacognitive knowledge with self-regulated learning. Zohar and Dori (2012) discuss how metacognitive instruction can be embedded in subject teaching.

Flavell, J. H. (1979)

Flavell (1979) defined metacognition, a key idea in learning. His framework includes knowledge, regulation, and experience. Teachers understand and use metacognition (Flavell, 1979) for learners.

Metacognition and Self-Regulated Learning: Guidance Report View guidance ↗
Policy guidance

Education Endowment Foundation (2018)

The EEF's work gives seven recommendations for teachers, plus classroom examples. They rate the strength of research behind each. Metacognition and self-regulation approaches have evidence across age groups and subjects (EEF). Implementing them well is associated with an average eight months of additional progress in the current EEF toolkit estimate.

Bjork et al. (2013) suggest effective learning techniques. Cognitive psychology provides helpful strategies. Dunlosky et al. (2013) found techniques improve learner results. Roediger and Karpicke (2006) show retrieval practice helps learners remember longer.

Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., and Willingham, D. T. (2013)

Dunlosky et al. (2013) rated ten study techniques, finding rereading and highlighting ineffective. Poor learner self-assessment explained these techniques' low scores. Retrieval practice and elaboration scored highly (Dunlosky et al., 2013). Teachers can use this to support effective learner study skills.

Becoming a Self-Regulated Learner: An Overview View study ↗
Peer reviewed

Zimmerman, B. J. (2002)

Zimmerman (2002) described forethought, performance and self-reflection in self-regulated learning. This connects thinking about thinking to learners' motivation, self-belief and self-evaluation. Keep the EEF evidence separate: it supports explicit teaching of metacognitive and self-regulated learning strategies rather than a generic growth-mindset claim.

Assessing Metacognitive Awareness View study ↗
Peer reviewed

Schraw, G. and Dennison, R. S. (1994)

Schraw and Dennison (1994) developed the Metacognitive Awareness Inventory around knowledge of cognition and regulation of cognition. Teachers can use this distinction to notice whether pupils know strategies, monitor their use and evaluate whether a strategy helped.

This week, choose one lesson where learners complete an extended task. Before they begin, give them two minutes to write down what they already know about the topic. Then ask what the task is asking them to do and which strategy they plan to use.

At the end, ask them to note what was harder than expected and what they would do differently next time. Collect the planning slips and compare them with the finished work. The gap between the plan and the final work will tell you more about their metacognitive regulation than any end-of-unit assessment.

‍

‍

‍

Frequently Asked Questions

‍

Metacognition Across Subjects

Metacognition can support learning in different subjects when the strategies fit the curriculum task. The current Education Endowment Foundation toolkit notes better outcomes in literacy, maths and science. Even so, teachers should adapt prompts to the learner and the subject. They should not claim one universal effect for every group.

Metacognitive reading strategies help learners check their understanding. They pause to summarise, question and predict; routines such as See, Think, Wonder and Think, Pair, Share can support talk when they are taught with a clear purpose and success criteria.

Thinking frameworks help learners, especially when they use digital tools. Higgins et al. (2004) found that regular practice improves self-regulation skills. Use these strategies often and consistently to strengthen learner metacognition.