Setting Common Metacognition Expectations for Learning with Your Students

by Patrick Cunningham, Ph.D., Rose-Hulman Institute of Technology

We know that students’ prior subject knowledge impacts their learning in our courses. Many instructors even give prior knowledge assessments at the start of a term and use the results to tailor their instruction. But have you ever considered the impact of students’ prior knowledge and experiences with learning on their approaches to learning in your course? It is important for us to recognize that our students are individuals with different expectations and learning preferences. Encouraging our students’ metacognitive awareness and growth can empower them to target their own learning needs and establish common aims for learning.

image of target with four colored arrows pointed at the center

Among other things, our students often come to us with having experienced academic success using memorization and pattern matching approaches to material, i.e., rehearsal strategies. Because they have practiced these approaches over time and have gotten good grades in prior courses or academic levels, these strategies are firmly fixed in their learning repertoire and are their go-to strategies. Further, when they get stressed academically, they spend more time employing these strategies – they want more examples, they re-read and highlight notes, they “go-over” solutions to old exams, they memorize equations for special cases, and more. And many of us did too, when we were in their shoes.

However, rehearsal strategies only result in shorter-term memory of concepts and surface-level understanding. In order to build more durable memory of concepts and deeper understanding, more effortful strategies are needed. Recognizing this and doing something about it is metacognitive activity – knowing about how we process information and making intentional choices to regulate our learning and learning approaches. One way to engage students in building such metacognitive self-awareness and set common expectations for learning in your course starts with a simple question,

‘What does it mean to learn something?”

I often ask this at the start of a course. In an earlier post, Helping Students Feel Responsible for Their Learning, I introduced students’ common responses. Learning something, they say, means being able to apply it or explain it. With some further prompting we get to applying concepts to real situations and explaining material to a range of people, from family member to bosses, to cross-functional design teams. These are great operational definitions of learning, and I affirm my students for coming up with them.

Then I go a step further, explaining how transferring to new applications and explaining to a wide range of audiences requires a richly interconnected knowledge framework. For our knowledge to be useful and available, it must be integrated with what we already know.

So, I tell my students, in this class we will be engaging in activities to connect and organize our knowledge. I also try to prepare my students for doing this, acknowledging it will likely be different than what they are used to. In my engineering courses students love to see and work more and more example problems – i.e., rehearsal. Examples are good to a point, particularly as you engage a new topic, but we should be moving beyond just working and referencing examples as we progress in our learning. Engaging in this discussion about learning helps make my intentions clear.

I let my students know that as we engage with the material differently it will feel effortful, even hard at times. For example, I ask my students to come up with and explore variations on an example after we have solved it. A good extension is to have pairs working different variations explain their work to each other. Other times I provide a solution with errors and ask students to find them and take turns explaining their thinking to a neighbor. In this effortful processing, they are building connections. My aim is to grow my students’ metacognitive knowledge by expanding their repertoire of learning strategies and lowering the ‘activation energy’ to using these strategies on their own. It is difficult to try something new when there is so much history behind our habitual approaches.

Another reason I like this opening discussion, is that it welcomes opportunities for metacognitive dialogue and ongoing conversations about metacognition. I have been known to stop class for a “meta-moment” where we take time to become collectively more self-aware, recognizing growth or monitoring our level of understanding. The discussion about what it means to learn something also sets a new foundation and changes conversations about exam, quiz, and homework preparations and performance. You might ask, “How did you know you knew the material?” Instead of suggesting “working harder” or “studying more”, we can talk meaningfully about the context and choices and how effective or ineffective they were.

Such metacognitive self-examination can be challenging for students and even a little uncomfortable, especially if they exhibit more of a fixed mindset toward learning. It may challenge their sense of self, their identity. It is vital to recognize this. Some students may exhibit resistance to the conversation or to the active and constructive pedagogies you employ. Such resistance is challenging, and we must be careful with our responses. Depersonalizing the conversation by focusing on the context and choices can make it feel less threatening. For example, if a student only studied the night or two before an exam, instead of thinking they are lazy or don’t care about learning, we can acknowledge the challenge of managing competing priorities and ask them what they could choose to do differently next time. We need to be careful not to assume too much, e.g., a student is lazy. Questions can help us understand our students better and promote student self-awareness. For more on this approach to addressing student resistance see my post on Addressing Student Resistance to Engaging in their Metacognitive Development.

Students’ prior learning experiences impact how they approach learning in specific courses. Engaging students early in a metacognitive discussion can help develop a common set of expectations for learning in your course, clarifying your intentions. It also can open doors for metacognitive dialogue with our students; one-on-one, in groups, or as a class. It welcomes metacognition as a relevant topic into the course. However, as we engage in these discussions, we must be sensitive to our students, respectfully and gently nudging their metacognitive growth. Remember, this is hard work and it was (and often still is) hard for us too!

Acknowledgements This blog post is based upon metacognition research supported by the National Science Foundation under Grant Nos. 1433757 & 1433645. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.

Distributed Metacognition: Are Two Heads Better Than One—Or Does it Even Exist?

by Aaron S. Richmond

Metropolitan State University of Denver

In many of the wonderful blog posts in Improve with Metacognition, scholars around the globe have described various teaching techniques and strategies to improve metacognition in our students. Many of these techniques require students to openly describe their learning behaviors with the hopes that they will become more metacognitively aware. For example, asking students how, where, and when they study, to reflect on the use of the strategies, and how they can improve their study strategies. Many of these examples are executed in a class setting and even sometimes students are asked to share their strategies with one another and discuss how these strategies work, when they work, when they don’t, etc. In such cases when students are sharing their beliefs about metacognition (e.g., learning strategies) we know that students benefit by improving their own metacognition through this process, but is it possible that they are improving the overall level of the group or class metacognition? Meaning, is it possible that there is more than just an individual metacognitive process occurring—is it possible that there is some form of distributed metacognition occurring across the students that is shared?

What is Distributed Metacognition?

Little is known about the concept of distributed metacognition (Chiu & Kuo, 2009; Wecker, & Fischer, 2007, July). In fact, in a Google Scholar search, there are only 38 results with the exact phrase “distributed metacognition”. In this limited research there is no clear operational definition of distributed metacognition. Therefore, I am interested in understanding and have a discussion with you all about the concept of distributed metacognition. From what I gather, it is not the spreading of metacognition over time (akin to distributed practice, spacing, or studying). Nor am I referring to what Philip Beaman (2016) referred to in the context of machine learning and human distraction in his IwM blog. However, could it be that distributed metacognition is the ability of two or more individuals to first talk and discuss their personal metacognition (the good, the bad, and the ugly) and to then to use these metacognitive strategies in a distributed manner (i.e., the group discusses and uses a strategy as a group)? Furthermore, Chiu and Kuo’s (2009) definition of social metacognition may be akin to distributed metacognition. Although they have no empirical evidence, they suggest that metacognitive tasks can be distributed across group members and thus they engage in “social metacognition”. For instance, in task management, students can simultaneously evaluate and monitor, regulate others to reduce mistakes and distraction, and divide and conquer to focus on subsets of the problem. Finally, in discussion with John Draeger (an IwM co-creater), he asked whether distributed metacognition was “…something over and above collaborative learning experiences that involve collective learning about metacognition and collectively practicing the skill?” After giving it some thought, my answer is, “I think so.” As such, let me try to give an example to illustrate whether distributed metacognition exists and how we may define it.

Using IF-AT’s Collaboratively

I have written on the utility of using immediate feedback assessment techniques (IF-AT) as a metacognitive tool for assessment. I often use IF-AT in a team-based and collaborative learning way. I have students get into groups or dyads and discuss and debate 1 or 2 questions on the assessment. They then scratch off their answer and see how they do. Regardless of whether they were correct or not, I have students discuss, debate, and even argue why they were so adamant about their answer as an individual and as a group. I then have students then discuss, debate, and answer two more questions with one another. They have to, as a group, come up with strategies for monitoring their performance, steps to solve the problem, etc. They repeat this process until the quiz is finished. When my students are doing this IF-AT process, I find (I know introspection is not the best science) that they become so intrigued by other students’ metacognitive processes, that they often slightly modify their own metacognitive processes/strategies AND collectively come up with strategies to solve the problems.

So, what is going on here? Are students just listening to other students’ metacognitive and epistemological beliefs and choosing to either internalize or ignore these beliefs? Or in contrast, when there is a group task at hand, do students share (i.e., distribute) the metacognitive strategy that they learned through the group process and then use it collectively? For example, when students perform activities like dividing tasks and assigning them to others (i.e., resource demand and monitoring), regulating others’ errors or recognize correct answers (i.e., monitoring) within the group— would these behaviors count as distributed metacognition? Is it possible that in these more collaborative situations, the students are not only engaging in their own internal metacognition, but that they are also engaging in a collective distributed cognition among the group used in a collective manner? That is, in the IF-AT activity example, students may be both becoming more metacognitively aware, changing their metacognitive beliefs, and experimenting with different strategies—on an individual level—AND they may also have a meta-strategy that exists among the group members (distributed metacognition) that they then use to answer the quiz questions and become more effective and successful at completing the task.

Currently (haha), I am leaning towards the latter. I think that the students might be engaging in both individual and distributed metacognition in part because of an article in the Proceedings of the Annual Meeting of the Cognitive Science Societyby Christopher Andersen (2003). Andersen found that when students worked in pairs to solve two science tasks, that over time, students who were in pairs rather than working individually made more valid inferences (correct judgments and conclusion about the task) than when they worked alone. Specifically, on the first trial of solving a problem, the dyads use relatively ineffective strategies, on the second trial they expanded and adapted their use of effective strategies, and by the third trial, the dyad expanded even more effective strategies. Andersen (2003) concluded that the students were collectively consolidating their metacognitive strategies. Meaning, when working collaboratively, students employed more effective metacognitive strategies that led to solving the problem correctly. Although this is only one study, it provides a hint that distributed metacognition may exist.

Tentative Conclusions

So, where does this leave us? As almost always, I am awash with questions. I have more questions than answers. Thus, what do you think? As defined, do you think that distributed metacognition exists? If not, how would you describe what is going on when students share their metacognitive strategies and then employ metacognitive strategies in a group setting? Is this situation just a product of collaborative or cooperative learning?

If you do believe distributed metacognition exists, how do we measure it? How do we create instructional methods that may increase it? Again, I am full of questions and my mind is reeling about this topic, and I would love to hear from you to know your thoughts and opinions.

References

Andersen, C. (2003, January). Distributed metacognition during peer collaboration. In Proceedings of the Annual Meeting of the Cognitive Science Society (Vol. 25, No. 25).

Beaman, P. (2016, May 14^th). Distributed metacognition: Insights from machine learning and human distraction. Retrieved from https://www.improvewithmetacognition.com/distributed-metacognition-insights-machine-learning-human-distraction/

Chiu, M. M., & Kuo, W. S. (2009). From metacognition to social metacognition: Similarities, differences, and learning. Journal of Educational Research, 3(4), 1-19. Retrieved from https://www.researchgate.net/profile/Ming_Chiu/publication/288305672_Social_metacognition_in_groups_Benefits_difficulties_learning_and_teaching/links/58436dba08ae2d217563816b/Social-metacognition-in-groups-Benefits-difficulties-learning-and-teaching.pdf

Richmond, A. S. (2017, February 22nd). Scratch and win or scratch and lose? Immediate feedback assessment technique. Retrieved from https://www.improvewithmetacognition.com/scratch-win-scratch-lose-immediate-feedback-assessment-technique/

Wecker, C., & Fischer, F. (2007, July). Fading scripts in computer-supported collaborative learning: The role of distributed monitoring. In Proceedings of the 8th international conference on Computer supported collaborative learning (pp. 764-772).

How Metacognition Helps Develop a New Skill

by Roman Taraban, Ph.D., Texas Tech University

Metacognition is often described in terms of its general utility for monitoring cognitive processes and regulating information processing and behavior. Within memory research, metacognition is concerned with assuring the encoding, retention, and retrieval of information. A sense of knowing-you-know is captured in tip-of-the-tongue phenomena. Estimating what you know through studying is captured by judgments of learning. In everyday reading, monitoring themes and connections between ideas in a reading passage might arouse metacognitive awareness that you do not understand a passage that you are reading, and so you deliberately take steps to repair comprehension. Overall, research shows that metacognition can be an effective aid in these common situations involving memory, learning, and comprehension (Dunlosky & Metcalfe, 2008).

image from https://www.champagnecollaborations.com/keepingitreal/keeoing-it-real-getting-started

But what about new situations? If you are suddenly struck with a great idea, can metacognition help? If you want to learn a new skill, how does metacognition come into play? Often, we want to develop fluency, we want to accurately and quickly solve problems. The classic model of skill development proposed by Fitts and Posner (1967) did not explicitly incorporate metacognition into the process. A recent model by Chein and Schneider (2012), however, does give metacognition a prominent role. In this blog, I will review the Fitts and Posner model, introduce the Chein and Schneider model, and suggest ways that the latter model can inform learning and development.

In Fitts and Posner’s (1967) classic description of the development of skilled performance there are three overlapping phases:

Initially, facts and rules for a task are encoded in declarative memory, i.e., the part of memory that stores information.
The person then begins practicing the task, which initiates proceduralization (i.e., encoding the action sequences into procedural memory), which is that part of memory dedicated to action sequences. Errors are eliminated during this phase and performance becomes smooth. This phase is conscious and effortful and gradually shifts into the final phase.
As practice continues, the action sequence, carried out by procedural memory, becomes automatic and does not draw heavily on cognitive resources.

An example of this sequence is navigating from point A to point B, like from your home to your office. Initially, the process depends on finding streets and paying attention to where you are at any given time, correcting for wrong turns, and other details. After many trials, you leave home and get to the office without a great deal of effort or awareness. Details that are not critical to performance will fall out of attention. For instance, you might forget the names of minor streets as they are no longer necessary for you to find your way. Another more academic example of Fitts and Posner includes learning how to solve math problems (Tenison & Anderson, 2016). In math problems, for instance, retrieval of relevant facts from declarative memory and calculation via procedural memory become accurate and automatic along with speed-up of processing.

Chein and Schneider (2012) present an extension of the Fitts and Posner model in their account of the changes that take place from the outset of learning a new task to the point where performance becomes automatic. What is distinctive about their model is how they describe metacognition. Metacognition, the first stage of skill development, “guides the establishment of new routines” (p. 78) through “task preparation” (p. 80) and “task sequencing and initiation” (p. 79). “[T]he metacognitive system aids the learner in the establishing the strategies and behavioral routines that support the execution of the task” (p. 79). Chein and Schneider suggest that the role of metacognition could go deeper and become a characteristic pattern of a person’s thoughts and behaviors: “We speculate that individuals who possess a strong ability to perform in novel contexts may have an especially well-developed metacognitive system which allows them to rapidly acquire new behavioral routines and to consider the likely effectiveness of alternative learning strategies (e.g., rote rehearsal vs. generating explanations to oneself; Chi, 2000).”

In the Chein and Schneider model, metacognition is the initiator and the organizer. Metacognitive processing recruits and organizes the resources necessary to succeed at learning a task. These could be cognitive resources, physical resources, and people resources. If, for example, I want to learn to code in Java, I should consider what I need to succeed, which might include YouTube tutorials, a MOOC, a tutor, a time-management plan, and so on. Monitoring and regulating the cognitive processes that follow getting things set up are also part of the work of metacognition, as originally conceived by Flavell (1979). However, Chein and Schneider emphasize the importance of getting the bigger picture right at the outset. In other words, metacognition can work as a planning tool. We tend to fall into thinking of metacognition as a guide for when things go awry. While we know that it can be helpful in setting learning goals so that we can track progress towards those goals and resources to help us achieve them, we may fall into thinking of metacognition as a “check-in” when things go wrong. Of course, metacognition can be that too, but metacognition can be helpful on the front end, especially when it comes to longer-term, challenging, and demanding goals that we set for ourselves. Often, success depends on developing and following a multi-faceted and longer-term plan of learning and development.

In summary, the significant contribution to our understanding of metacognition that Chein and Schneider (2012) make is that metacognitive processing is responsible for setting up the initial goals and resources as a person confronts a new task. With effective configuration of learning at this stage and sufficient practice, performance will become fluent, fast, and relatively free of error. The Chein and Schneider model suggests that learning and practice should be preceded by thoughtful reflection on the resources needed to succeed in the learning task and garnering and organizing those resources at the outset. Metacognition as initiator and organizer sets the person off on a path of successful learning.

References

Chein, J. M., & Schneider, W. (2012). The brain’s learning and control architecture. Current Directions in Psychological Science, 21, 78-84.

Chi, M. T. (2000). Self-explaining expository texts: The dual processes of generating inferences and repairing mental models. In R. Glaser (Ed.), Advances in instructional psychology, (Vol. 5), pp. 161-238. Mahwah, NJ: Erlbaum.

Dunlosky, J., & Metcalfe, J. (2008). Metacognition. SAGE, Los Angeles

Fitts, P. M., & Posner, M. I. (1967). Human performance. Belmont, CA: Brooks/Cole.

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

Tenison, C., & Anderson, J. R. (2016). Modeling the distinct phases of skill acquisition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42(5), 749-767.