top of page

How the EEF gets metacognition and self-regulation wrong - and why it matters

This blog post features an important, adapted extract from Fear is the Mind Killer. I discussed this in a recent podcast with Christian Bokhove, one of the authors of the EEF evidence review on metacognition and self-regulation. You can find video and audio versions of our (fascinating!) interview here.

Introducing the 'frog model' of self-regulated learning [1]

There has been considerable interest in metacognition and self-regulation since the Education Endowment Foundation (EEF) declared in 2011 that they provide ‘high impact for very low cost, based on extensive evidence’. [2]

However, teachers are not always clear about what ‘metacognition and self-regulation’ means, or what these ideas look like in the classroom. This is not surprising, because educational researchers aren’t always clear about what it means, either.

Dinsmore et al. (2008) reviewed over 250 studies in an attempt to determine the ‘core meaning of metacognition, self-regulation and self-regulated learning, as well as where these constructs converge and diverge’. [3] This review found that only 49% of the studies provided explicit definitions, and that where this did happen, there were considerable differences, as well as areas of overlap, between the three constructs. Dale Schunk (2008) summed up this exasperating situation admirably:

'These definitions have become diluted to the point where today we ask such questions as: Is metacognition part of self-regulation? Is self-regulated learning part of self-regulation? Is self-regulation more environmentally sensitive than metacognition, which is more of a personal factor?’ [4]

In 2018, the EEF published a guidance report entitled Metacognition and Self-regulated Learning, in an attempt to simplify this complex field for busy teachers. [5] The guidance report outlines seven key recommendations for schools. These are great, and I would urge all teachers and school leaders to read, digest and implement them forthwith.

However – and bear with me if this seems pedantic, but I really do think it’s important – there are a number of problems with the way in which the EEF defines its terms. If we're going to get to grips with how to teach children to become confident, independent learners, we need to understand the theory (the why) as well as the practice (the what). In this way, teachers can implement these powerful ideas from first principles, rather than taking recommendations off the shelf without a clear understanding of the underlying concepts.

The EEF defines self-regulated learning as a broad umbrella concept comprising three sub-components – cognition, metacognition and motivation (Figure 1).

Figure 1. The EEF model of self-regulation/self-regulated learning [7]

They define these terms as follows:

Cognition – the mental process involved in knowing, understanding, and learning
Metacognition - often defined as ​‘learning to learn’
Motivation – willingness to engage our metacognitive and cognitive skills [6]

There are three significant problems with the way in which the EEF defines these terms.

First, as we will see, this is not how metacognition and self-regulation are understood by the people who understand them most. I have read hundreds of articles on metacognition and I have never seen it defined as 'learning to learn'. The EEF definition therefore introduces confusion into an already complex arena. [8]

Second, the EEF model is overly concerned with cognition, and overlooks the importance of feelings - both physical and emotional. Learning how to to monitor and control our feelings – and therefore our behaviours – is really the bedrock of self-regulated learning. This goes way beyond motivation - important though motivation undoubtedly is.

And third, the EEF uses the terms self-regulation and self-regulated learning interchangeably. In their guidance document, there are 42 references to self-regulated learning and 19 references to self-regulation, and no attempt is made to differentiate between the two. This is a problem because to understand how these concepts help define one another is actually quite illuminating. As Alexander (2008) put it: “How better to comprehend the nature of metacognition… than to ponder its associations with self-regulation or self-regulated learning?”. [9]

How indeed? Let’s consider each in turn.


Metacognition is often referred to simply as ‘thinking about thinking’, which often prompts hilarious gags about ‘thinking about thinking about thinking’. [10] However, when John Flavell coined the word in 1976, he defined it as a complex, dynamic process that goes far beyond merely ‘thinking about thinking’.

To be precise, Flavell defined metacognition as ‘the active monitoring and consequent regulation and orchestration of [thought] processes… usually in the service of some concrete goal or objective’. [11] In 1979, Flavell developed his thinking further in a short paper called Metacognition and cognitive monitoring, which remains the best model we have (Figure 2). [12]

Figure 2. Flavell’s model of metacognition and cognitive modelling (Source: Impact)

To break this down a little, Flavell, a developmental psychologist, suggested that we learn to control our thinking by monitoring what we know about people (self and others), tasks and strategies. He proposed that this metacognitive knowledge grows through experience, setting goals and identifying and using strategies to achieve those goals. All of these components interact with one another, and through these interactions we develop metacognitive skills and further our knowledge, improving our ability to independently achieve our goals in the future.

You can see why the EEF wanted to simplify things: Flavell’s model has great explanatory power, but it is too complex to be easily called to mind in the context of a busy classroom. However, metacognition can be simplified without distorting the message too badly.

For example, Chris Watkins – a giant of the self-regulated learning movement [13] – defined metacognition succinctly as ‘awareness of thinking processes, and ‘executive control’ of such processes’. [14] Here, we propose an even simpler definition:

Metacognition is monitoring and controlling your thought processes


Our understanding of self-regulation is largely based on the work of the psychologist Albert Bandura in the 1970s and 80s. In contrast to the cognitive, thought-based world of metacognition, Bandura viewed self-regulation as the process of influencing the external environment through our feelings and behaviours. [15]

The language used to describe self-regulation is often strikingly similar to that used to describe metacognition, and in a sense the two can be seen as mirror images of one another. In their aforementioned review, Dinsmore et al. (2008) found that there was significant overlap in the language researchers use to define the two terms, with two words cropping up far more than any others: monitor and control.

Echoing Bandura, Dinsmore et al. concluded that there is ‘a clear cognitive orientation for metacognition, while self-regulation is as much concerned with human action as the thinking that engendered it’. [16]

The importance of monitoring and control is also emphasised by van Merriënboer and Kirschner (2018):

‘Monitoring is the term used to refer to the — metacognitive — thoughts learners have about their own learning. For example, learners who are reading a study text should monitor their level of comprehension of the text. Control refers to how learners respond to the environment or adapt their behaviour based on their thoughts. Thus, if comprehension monitoring leads a learner to the thought that a text is not yet well understood, (s)he may decide to restudy one or more parts of this text. Monitoring and control are closely linked to each other in one and the same learning cycle: One is of no use without the other.’ [17]

Here, van Merriënboer and Kirschner suggest that monitoring is primarily a metacognitive concern, while control relates to an individual’s behaviours. However, we take a slightly different view: we can monitor and control our thoughts, feelings (physical and emotional) and behaviours.

Our simplified definition of self-regulation is therefore as follows:

Self-regulation is monitoring and controlling your feelings and behaviours

Self-regulated learning

Thoughts, feelings, behaviours… clearly, metacognition and self-regulation are broad concepts that extend far beyond academic learning. You can think about anything - what's for dinner, the fact that your cat needs to go to the vet, why there's so much month left at the end of the money - and so on. Likewise, feelings and behaviours are vast categories that extend far beyond learning. As Fox and Riconscente (2008) put it: ‘understanding metacognition and self-regulation… requires situating them within the broad context of all activities for humans of all ages and points of development’. [18]

Following the publication of Bandura's classic 1986 work Social Foundations of Thought and Action, the concept of self-regulation was increasingly applied to the process of learning. [19] This led to the development of a new term, ‘self-regulated learning’.

Schunk (2008) describes self-regulated learning as ‘the process whereby students activate and sustain cognitions and behaviours systematically oriented toward the attainment of their learning goals’. [20] Here, we propose the following, simplified definition:

Self-regulated learning is the application of metacognition and self-regulation to learning

To recap:

  • Metacognition is monitoring and controlling your thought processes;

  • Self-regulation is monitoring and controlling your feelings and behaviours; and

  • Self-regulated learning is the application of metacognition and self-regulation to learning (Figure 3).

Figure 3. The 'frog model' of self-regulated learning

As you can probably see, when we created this visual, it accidentally came out looking a bit froggy - hence the name.

As with all models, this is a simplified version of reality. For example, there are a number of additional factors that influence the extent to which a student is able or willing to regulate their own learning. This includes motivation, as the EEF’s model recognises (to what extent are students motivated to learn, and is this intrinsic or extrinsic?). But there are additional factors to consider, such as autonomy (do the students have any choice about whether or how to engage in their learning, or are the expectations imposed by the teacher, or a looming deadline?), self-efficacy (to what extent do they believe they can learn effectively within this domain?), and so on.

For this reason, in the Activate resource, we identified five 'enabling factors' that support self-regulated learning (Figure 4).

Figure 4. The 'frog model' of SRL with 5 enabling factors

As ever, context is king. However, we believe that the 'frog model' of self-regulated learning outlined above is more helpful than that proposed by the EEF, for three reasons.

First, our model does not introduce new concepts or definitions into an already overcrowded field. Rather, it simplifies established definitions of metacognition, self-regulated and self-regulated learning, and explains how they relate to one another.

Second, it makes clear that metacognition and self-regulation are essentially mirror images of one another – the former relating to the monitoring and control of thoughts, the latter relating to the monitoring and control of feelings and behaviours. This key insight - that both metacognition and self-regulation are essentially comprised of two processes – monitoring and control – is of great practical use when thinking about how these ideas play out in the classroom.

And third, it clarifies the difference between self-regulation and self-regulated learning. In so doing, it recognises the importance of emotional and physical self-regulation as prerequisites to self-regulated learning, which are critical even in early childhood. [21]

This is all well and good, but what does it look like in practice?

In schools, teachers set the agenda for what needs to be learned, and how, and when. We set deadlines for the pupils to meet; we remind them of those deadlines regularly; and, if it looks like a pupil is not going to meet that deadline, we swoop in and ‘intervene’ – especially as they approach GCSE. We organise after-school catch-up sessions, we phone their parents and carers, we send reminders via their tutor… we do whatever it takes to get them over the line. [22]

All of this is done with the best of intentions, mindful of the need to use our time efficiently and to maximise the students’ chances of success. But the question almost asks itself: how can we expect children to learn how to self-regulate in such a top-down, micro-managed environment?

The answer is obvious, although the implications for what happens in schools is far-reaching: in order for children to learn how to regulate their thoughts, feelings and learning behaviours, teachers need to occasionally take a step back, to find out (and to allow the children to discover) what they can and cannot do by themselves.

This process of stepping back can be done in a measured way, for just a part of a lesson – or, we can really ‘throw them in at the deep end’ and set them a task that lasts for several weeks.

To extend the metaphor: clearly, we do not want our pupils to drown, or to flail around fruitlessly for lesson after lesson. But nor should we throw them with a lifeboat at the first sign of struggle. Instead, we might toss them a branch – something to help them keep their head above water until they can survey their surroundings and build a raft of their own. OK, we’ve definitely taken the metaphor too far now. You get the idea.

In practical terms, there are three parts to this process.

First, set them a challenging task, with the expectation that they will produce something of high quality. Something that takes a half-term to complete. Something they have never done before; something they may feel is currently beyond them; something that requires them to stretch out into new territory – just beyond the edge of their comfort zone. (See my recent podcast with Ron Berger for some examples of what this looks like in practice)/.

Next, we provide them with the support and guidance they need to get started, or to nudge them back on track when they get stuck.

And finally, we withdraw that support over time, until they are able to achieve success without constant supervision.

This is closely related to Jerome Bruner’s notion of instructional scaffolding, the idea that teachers should provide children with the guidance and support they need to learn something, and then gradually withdraw that support as the child grows in confidence or proficiency. [23]

Scaffolding is a useful framework for most forms of instruction. But when the aim is self-regulated learning, we need to take the scaffolding idea even further because if we provide guidance and support too readily – at the first sign of difficulty, or in response to the first request for help – we can breed dependence.

When the goal is self-regulated learning, therefore, the question becomes: what is the minimum scaffolding we need to have in place? When we ‘step back’, should we remain one step behind the child, or ten? This process of stepping back needs to be done judiciously – choosing carefully when to do so, and why, and for how long. It’s probably not a good idea to trial this a few weeks before the GCSE exam.

In our view, Key Stage 2 and 3 is the best time to do this work, when children are aged roughly between 8 and 14. The need to step back is one of the reasons why the Learner Effectiveness Programme entails discrete lessons, as well as being embedded throughout the curriculum. A child cannot expect to become a stronger swimmer without practising swimming repeatedly.

Similarly, if we want children to become self-regulated learners, we have to provide them with the time and space in which to develop and practice the knowledge, skills, habits and dispositions that underpin self-regulated learning.

To be clear, ‘stepping back’ does not mean indulging in some uber-permissive, child-centred experiment in unguided discovery learning, hoping that they will figure out for themselves how to manage their feelings and behaviours.

We can teach children how to self-regulate using the tried and tested, traditional methods that we use to teach most things. We explain it. We model what it looks like. We deconstruct it and make sure they understand all the moving parts. We co-construct it with them (co-regulating in this case). We provide feedback and give them time to respond to it and act on it. And we gradually withdraw our support over time.

Progressive ends through traditional means, you might say.

Do you agree that the 'frog model' of self-regulated learning is clearer and more helpful than the EEF definition? If so, why? If not, why not? Please feel free to share your thoughts below!


[1] This section is based on Mannion, J. (2020) Metacognition, self-regulation and self-regulated learning: What’s the difference? Impact, the Journal of the Chartered College of Teaching, 8 (Spring), p66-69.

[2] Higgins, S., Kokotsaki, D. & Coe, R.J. (2011). Toolkit of Strategies to Improve Learning: Summary for Schools Spending the Pupil Premium. Sutton Trust. Available at:

[3] Dinsmore, D.L., Alexander, P.A., Loughlin, S.M. (2008). Focusing the conceptual lens on metacognition, self-regulation, and self-regulated learning. Educational Psychology Review, 20(4), p392.

[4] Schunk, D.H. (2008). Metacognition, self-regulation, and self-regulated learning: Research recommendations. Educational Psychology Review, 20(4), p465.

[5] Education Endowment Foundation (2018). Metacognition and self-regulated learning: Guidance report. London: EEF, p4.

[7] Muijs, D. & Bokhove, C. (2020). Metacognition and Self-Regulation: Evidence Review. London: Education Endowment Foundation.

[8] The EEF’s definition was used by Schraw et al. (2006) – however, this is an exception, rather than the norm. See Schraw, G., Crippen, K.J., Hartley, K. (2006) Promoting Self-Regulation in Science Education: Metacognition as Part of a Broader Perspective on Learning. Research in Science Education, 36, 111–139.

[9] Alexander, P. (2008). Why This and Why Now? Introduction to the Special Issue on Metacognition, Self-Regulation, and Self-Regulated Learning. Educational Psychology Review, 20(4), p370.

[10] See also ‘Learning to learn to learn to learn’.

[11] Flavell, J. H. (1976). Metacognitive aspects of problem solving. In L. B. Resnick (Ed.), The nature of intelligence (pp. 231-235). Hillsdale, NJ: Lawrence Erlbaum, p232.

[12] Flavell, J.H. (1979). Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist, 34(10), 906-911.

[13] To peruse Chris’s phenomenal output on Learning to Learn and other topics over the years, visit

[14] Watkins, C. (2001). Learning about learning enhances performance. London, Institute of Education National School Improvement Network (Research Matters series No 13), p1.

[15] Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and Human Decision Processes, 50, 248-287.

[16] Dinsmore, D.L., Alexander, P.A., Loughlin, S.M. (2008). Focusing the conceptual lens on metacognition, self-regulation, and self-regulated learning. Educational Psychology Review, 20(4), p405.

[17] van Merriënboer, J. J. G., Kirschner, P. A. (2018). Ten steps to complex learning (3rd ed.). New York, NY: Routledge.

[18] Fox, E., Riconscente, M. (2008). Metacognition and Self-Regulation in James, Piaget, and Vygotsky. Educational Psychology Review, 20(4), p374.

[19] Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.

[20] Schunk, D.H. (2008). Metacognition, self-regulation, and self-regulated learning: Research recommendations. Educational Psychology Review, 20(4), p465.

[21] Whitebread D. & Basilio M. (2012). The emergence and early development of self-regulation in young children. Profesorado: Revista de Curriculum y Formacion del Profesorado, 16, 16-33.

[22] It is worth noting that this is not typical in some countries. It is, however, very much the case in the UK – especially within the state secondary sector, and especially in Years 10 and 11.

[23] Bruner, J. S. (1978). The role of dialogue in language acquisition. In A. Sinclair, R., J. Jarvelle, and W. J.M. Levelt (eds.) The Child's Concept of Language. New York: Springer-Verlag.

[24] Claxton, G. (2012) Virtues of Uncertainty. Aeon magazine, September 17. Available at:

bottom of page