“We don’t learn from experience, but from reflecting on experience.” (John Dewey, although disputed)
I know the buzzwords in education have invaded us in the past few years and, hopefully, we can navigate media better so as to not get trapped in them. As fancy as they may sound, they are not far from what we have already been doing as teachers and I am not certain an inflation of edu-words helps us in our practice, on the contrary. However, since these words are being circulated often, I decided to discuss one, “metacognition”. My source is Handbook of Research on Learning and Instruction, more specifically chapter 10 (pages 197-219).
Historical roots
In the educational literature two terms are being used interchangeably despite having different conceptual roots and theoretical perspectives.
Metacognition theory: it originated from developmental psychology with Piaget (Inhelder & Piaget, 1958) and Flavell (1970) as progenitors. It focused on “reflective abstraction of new or existent cognitive structures”; simply put, a person’s thinking about own thinking/cognition.
Self-regulated learning (SRL): emerged from metacognition theory above and Bandura’s self-regulation theory. The emphasis is on the regulation of learning processes and outcomes (Dinsmore et al, 2008; Zimmerman, 1995). In other words, knowing what, how and when to apply a specific strategy to monitor, carry, and self-evaluate learning.
The line between the two, however, became fuzzy over time and researchers themselves plead for more clarity in conceptual and operational definitions.…………………………………………………………………………………………………………………
Why does this matter?
Wang, Haertel, and Walberg (1990) concluded from their literature review that metacognition is the most important predictor of learning performance.
Veenman , in an overview of studies with learners of different ages, performing different tasks in various domains (2008), estimated that metacognitive skillfulness accounted for 40% of variance in learning outcomes.
I think this important in three ways.
First, it gives us an idea of the complexity of factors at play in learning, beyond acquisition of knowledge. A student might very well consolidate their domain knowledge but it in the absence of these self-regulatory mechanisms and understanding of own thinking they might invest too much or less effort on certain tasks.
Secondly, because self-regulation systems take time to develop and often teachers become frustrated with what might appear as lack of understanding or skill. “How is it possible that after I taught … and had so many practice tasks, student A cannot succeed yet?” This apparent lack of knowledge or skill can be entirely dependent on SRL: the student might very well *have* the knowledge but doesn’t know when and how to retrieve and apply it.
A third reason is that two extreme teaching approaches can be detrimental: too much direct instruction and discovery learning. In the presence of continuous explicit teaching, where the teacher consistently guides the solving of a task (may it be the analysis of a literature piece or a mathematical problem) and there is a unidirectional assessment, students have little opportunity to develop these mechanisms. On the other hand, when students are continuously challenged to discover patterns, concepts, relationships and so forth, their self-evaluation and monitoring processes are overloaded and can be, often, incorrect.…………………………………………………………………………………………………………………
It gets complicated
Metacognitive knowledge refers to one’s knowledge about the interplay between three factors (Flavell, 1979):
- person
- task
- strategy
For instance, a learner may think that s/he (person characteristic) is not proficient in reading (task characteristic) and, therefore, that s/he should invest more effort in studying a textbook chapter (strategy characteristic). The reverse is true as well (investing too little effort in tasks one *thinks* is already proficient at).
It is easy to notice that metacognitive knowledge can be either correct or incorrect as learners underestimate or overestimate their competences (Veenman et al., 2006). A student may erroneously think that s/he needs only to read a chapter once in preparation for an exam despite repeated failures.
This self-knowledge may also prove resistant to change, especially when failure is misattributed to external causes (poor teachers, unsound exams).
Moreover, even correct metacognitive knowledge does not guarantee an adequate execution of appropriate strategies, as the learner may lack either the ability or the motivation to do so (this discrepancy between children’s knowledge and skill was found in several studies – Carr et al, 1995; Winne, 1996; Veenman, 2005).
In light of so many variables, researchers realized that there are 3 components to metacognition:
- metamemory (basically factual knowledge, knowing what)
- conditional knowledge (knowing when to apply a strategy)
- procedural knowledge (knowing how to apply it)
(See Flavell, 1970; Flavell & Wellman, 1977; Cavanaugh & Permutter, 1982; Nelson & Narens, 1990; Schraw & Moshman, 1995; Kuhn, 1999; Alexander & Jetton, 2000; Anderson & Schunn, 2000; Zohar & Ben-David, 2009).…………………………………………………………………………………………………………………
Where does metacognitive knowledge come from?
Students, and adults alike, gain and build this knowledge from three sources:
- personal belief system (that contains naïve theories and tacit ideas about their own cognition)
- judgments and feedback from other people
- metacognitive experiences (these are non-analytic, non-conscious inferential processes that are driven by affective factors, such as interest, disappointment, curiosity)
Metacognitive experiences are completely subjective by nature. For instance, while a task may have an externally defined objective level of difficulty or cognitive load (Sweller, 1994), the learner subjectively appreciates its difficulty based on self-perceived cognitive abilities as well as affective elements such as fear of failure (Efklides, 2006).
“These experiences affect performance DIRECTLY through time on task and effort expenditure.” (Veenman, 2005).…………………………………………………………………………………………………………………
What is the developmental timeline of metacognitive knowledge?
Piaget did not expect metacognition to develop before formal-operational thinking stage because”young children’s egocentrism prevents them from being able to introspect or treat their own thought processes as an object of thought” (Inhelder & Piaget, 1958).
Flavell and others discussed about “proto-metacognition” : a stage of rudimentary metacognition beginning to unfold at the ages of 4 -6. It is very basic in its nature and it is described as children’s acknowledgment that people may see different things, although they cannot handle the various perspectives people take. (see also Istomina, 1975; Kluwe, 1987; Demetriou & Efklides, 1990; Kuhn, 1999; Van Otterloo et al., 2004).
Other researchers observed elementary forms of planning, monitoring, and reflection when that task is appropriate to children’s level of understanding and interest (such as doll play, sorting out buttons by shape etc.). Whitebread et al. argued that earlier studies underestimated metacognitive processing in pre-school children because the assessment methods relied too much on children’s verbal abilities, which develop slower. (see also Larkin, 2006)
However, generally, the development of metacognitive skills is thought of as commencing at the age of 8-10 years (Berk, 2003; Veenman et al., 2006). As children develop knowledge and strategies, they expand their metacognition and school has a critical effect because it provides them with academic repertoires and more sophisticated tools as well as feedback on their learning. (see Alexander et al., 1995; Veenman, 2006; Van der Stel & Veenman, 2010)
What is interesting, however, is that intelligence does not affect the development of metacognition. (Veenman et al., 2004; Veenman & Spaans). I find this important because in teaching environments that emphasize either too much teacher instruction or student discovery, self-regulation can be hindered, albeit in different ways. This has implications on self-directed learning and teachers need to mindful how far they can go with independent tasks at different ages (see my arguments in another post I wrote, No, I Don’t Personalize Learning). Reversely, in settings where the teacher is the sole person in the room that gives feedback and orientates learning, self-regulation processes may be slowed down.
…………………………………………………………………………………………………………………
OK, I understand this meta-fluff. What about the teaching?
Even armed with this knowledge about metacognition, one can ask themselves, “Well, these self-regulation mechanisms are covert. We can only infer them from student performance (i.e. the student, independently, revises the problem s/he solved to identify the error) and reactions (i.e. curiosity about a task). How can I use this in my daily teaching?” Indeed, considering the factors that influence metacognition, we can be baffled. How can we influence someone’s belief system or their experience driven by affective elements?
Well, we can. It takes time and the development is incremental, strenuous at times. We need to take into account what actually helps humans produce this system of metacognitive processes.
Stage 1 – Cognitive stage
Describe the procedure with clarity. Verbalize what needs to be done, and in what order. This is basically guiding the acquisition of factual knowledge.
The conscious execution of the procedure at this stage explains why the initial acquisition of metacognitive skills through instruction or training requires extra effort, which initially may interfere with cognitive performance. (Schraw et al., 2006; Veenman et al., 2006).
Examples are from any domain: how to add fractions, how to use a punctuation mark, how to solve a problem in steps, think-alouds during reading etc.
Stage 2 – Associative stage
Simply put, it means practice plus self-monitoring structures. Allow students to practice, in different contexts and at different levels of difficulty. Help them monitor their learning with external prompts (rubrics, checklists, cues and questions; i.e. “Have I added…? Have I capitalized my sentences? etc.)
In this stage, verbal descriptions of the procedure are transformed into a procedural representation. It is the blurry line when knowledge becomes skill and can be retrieved, initially with errors of course, without effort. The execution of procedures “becomes faster and more accurate, requiring less effort.”
Stage 3 – Autonomous stage
Begin to remove scaffolding or other prompts. Students need independence at this point and what helped them in the previous stage may actually interfere with learning (also, see expertise reversal effect, Applying the Science of Learning in Education, 2014).
At this stage, “monitoring processes run in the background until an error or anomaly is detected.” (Brown, 1997; Butler & Winne, 1995; Reder &Shunn, 1996) Along the same line, “elements of the planning process may become automated, thus requiring less deliberate and conscious activity.” (Pressley et al., 1989)…………………………………………………………………………………………………………………
Now, one may say that these are simply good teaching strategies and they bring nothing new to the discussion. I partly disagree.
Firstly, it is because some of my readers might not be experienced teachers. We all are on a continuum of learning ourselves and what might be obvious to some can be a helpful guide for others.
Secondly, because it helps us tread more cautiously on extremes (direct teaching and discovery learning). Personally, I believe in balance and in a well-informed opinion on education. More often than not we form our pedagogical beliefs in the absence of research and that can impact student learning indirectly, through the strategies we choose. I always use research, suggestions from other teachers, and personal experience to “triangulate”, if you wish, what would make a good strategy.
Thirdly, it is because my own experience showed me how it is because of little things that student learning can change. Super innovative, one-time lessons are great but the impact does not last. I’ll give you a simple example.
I had a student who seemed brilliant in our Math Talks and different tasks: he could find patterns, noticed relationships, and could give proofs where others were unable to. However, in written tests he made a mess. The first three were a disaster: less than 30% of tasks were completed. Even in our daily math warm-up (6-a-day) that I created he would often make mistakes, despite having to deal with tasks that were of average difficulty. I then read research literature and tried to come up with my own strategy. What did I do? I simply gave him one test task at a time, glued on separate strips of paper. As soon as he finished, I would glue the next one and so on. The improvement? 90% of the tasks were completed correctly. Now I couldn’t have come up with this strategy without having previously been informed. I read about test anxiety, about attention deficit, and about sequencing of activities for different learners. Without this reading I would probably have been struggling a lot longer to figure out what was that I wasn’t doing properly. Later, I modified the strategy so he can become autonomous.
Moments, Snippets, Spirals 