Learning how to learn & teach
It’s been a ~decade of accelerating bidirectional communication between cognitive science and classroom teaching. And it seems that the field is ready for a leap forward. Some thoughts on the basis of recent selected key publications🧵⬇️
3 They review metacognitive barriers to implementation (e.g. misconceptions, greater effort) and suggest that to develop self-regulated learners we should plan means to promote conceptual change and drive behavior on the basis of a 4-component framework.➡️
4 The Know-Belief-Commitment-Planning framework by McDaniel and Einstein (2020) is designed on the basis of existing evidence to aid the application challenges and to promote students' self-regulated learning. The 4 components: ➡️ bit.ly/3Szb5Y6
5 Knowledge, Believe, Commitment, Planning (KBCP) illustrate how the development of self-regulating learning takes more than knowledge and/or motivation. They describe the evidence supporting each part, give an example, and suggest the framework as a basis for future research➡️
6 The KBCP model calls for a comparison with the IGTP model for effective teacher professional development, formulated by @DrSamSims@HFletcherWood and colleagues via @EducEndowFoundn (2021) as a basis for a systematic review and meta-analysis ➡️ bit.ly/3E7J3hJ
7 The IGTP model identifies 14 mechanisms from research on how people learn and change their practice and organizes them under 4 major categories:
8 They suggest that a balanced CPD program should include at least one mechanism from each category. This makes a lot of sense: highlighting major milestones of behavior change but also leaving room for adjustment and adaptations to audiences, contexts, and goals.
9 Thinking about the two frameworks and comparing them raises many intriguing questions. As I see it, such questions represent the next stage in the CogSci-Edu field. Note some key common features between the two models: ➡️
10
-Focus on behavior change
-Identify barriers to effective strategy use
-Identify key steps that span: declarative learning, metacognition, motivation, practice & habit formation
- Based on evidence and call for further research
And what are the differences?
As a last note, this is my recent attempt to depict the interplay between the basic learning process, the related research fields, and some goals. It may explain my excitement and maybe help in painting the gradually clarifying "big picture".
This is an amazing thread unpacking the model and much more... I wish I could edit and link:
Short 🧵on Learning and Memory in the brain:
'Neuroplasticity' is everywhere, but what do we really mean when we talk about the ever-changing brain?
Let's dive deeper than the buzzword and explore the evidence with a model.
2. Cognitive neuroscience uses simplified network models like this to demonstrate how learning & memory might work at the network level.
Nodes represent neurons, lines their connections (synapses), and the patterns - bits of our knowledge.
3. This model highlights two key features of neuroplasticity: 1) Existing nodes & connections can be inactive or reactivated. 2) Activating new patterns can sometimes forge new connections (but generally not new nodes).
1 How should we use Generative AI for Academic teaching?
The answer, imo, is in cognitive science, as the human learning process is both the goal and the limiting factor in this journey.
A🧵
#HigherED #GenAI #CogSci
2 How can GenAI be used in academic teaching? Which skills will become obsolete? Is academic teaching going to change completely?
So many questions as we are perplexed by the GenAI Stuns.
However, we have some powerful tools to think about it rationally:
3 The most important distinction is between experts and novices. For experts, GenAI is very helpful: you can use it sophisticatedly to save time and improve your work. You can evaluate when it is helpful and when it is not. But what does it take?
1/ This book and this app have convinced me, through theory and practice,
how important it is to include habit formation in every educational program or plan, at any level 🧵🤓
2/ We are naturally biased against investing in habit formation, as it is mostly unconscious and long-term.
We are way more easily convinced by logical reasoning for behavioural change:
setting goals, finding willpower, and boosting motivation, all seem compelling, but…
3/ They don’t work in the long-term unless we also invest in forming good sustainable habits.
It’s true for each of us when pursuing our goals like learning a language or exercising regularly, but it is even more crucial in educational settings:
1/ What is the role of errors in learning? And what is an error, really?
The “Derring Effect”- making errors deliberately to improve learning, is newly described by Wong & Lim (2022).
It triggered some thoughts around errors, nerd out with me 🤓🧵:
2/ First, the evidence: following studying a short academic text, a practice session that included making errors deliberately and then correcting them (in writing) was more effective, when measured with an application test than several carefully designed control groups:
3/ Deliberately erring and correcting key concepts, in writing, was more effective than:
- Copying the text and underlying key concepts
- Creating a concept map with the key concepts
- Generating synonyms for key concepts
(Wong & Lim, 2022)
Yesterday at @researchEDWarr I tried to make the point that the “love story" between cognitive science and education is exciting because it has reached the point where it’s not just about isolated “quality ingredients”, but about the entire “dish” and even a whole “meal”.
I’ve put 4 things on the table, 4 points that I find central and essential for making cog sci useful in education
1. It’s essential to acknowledge the limitations of the cognitive systems:
📌Attention and WM are limited in capacity, the bottleneck of processing.
📌LTM is “Blackboxy” – we don’t really know how it works, parts of it are unconscious and we are biased as a result
1/ How Predicting is different from guessing? and from practicing retrieval? All three strategies are based on generating a response based on semantic elaboration, but how do they differ, and why prediction is worth special attention as a standalone strategy? A thread:
2/ In this new paper @garvin_brod suggests that Prediction deserves the attention, of both teachers and researchers, as an effective strategy for learning, and I agree. Why?
3/ Prediction is different from guessing because it's based on relevant prior knowledge and is associated with higher confidence. This leads to curiosity and surprise if feedback reveals a violation. Surprise enhances focusing attention and encoding the new info as valuable.