These will be raw notes taken in real time and undergoing very little editing. They will be words from the speaker blended with my own thoughts as I process what is being said. While I will try to note the difference, I can't promise that will always happen. Don't hold a speaker responsible for anything I put here.
Keynote Address I: The Neurobiology of Learning to Learn: What We Think, We Become by André A. Fenton
I was 12 minutes late because navigating the NYC public transit system is challenging for the uninitiated.
The stuff he is explaining is far too complex with neurology for my notes to make any sense. So, these will be limited.
Complex systems - Showed a video of a large flock of birds in a big complex flight pattern.
Three rules:
- Keep flying.
- Don't bump into anything.
- Do whatever your seven nearest neighbors are doing.
A good analogy for memory is a large language learning model.
Deep learning neural networks don't remember things.
They store processes that make predictions that look like recollection.
We don't see and experience the world the way it actually is.
We see it the way our neuron's experiences cause us to interpret it. It's based on what we are expecting.
A mouse was exposed to a mild electric shock in a certain area.
90 minutes teaches it to avoid the area.
Mice with memory training avoided the area 30 days later (which is a long time in a mouse life).
It didn't store information; it changed how its brain was operating during that month.
Reactivating the right set of neurons in some form recreates the experience.
Large language models (AI) are built from the same models as our real neural networks. John Hopfield and Geoffrey Hinton won the Nobel Prize for the discoveries of memory networks because they are leading to AI networks.
Cognitive control is taxing. If you try to focus on one thing while ignoring something else (like saying the word "red" when you see this BLUE. But making it a little more difficult results in longer term learning because the work required changes the neural circuitry.
Keynote Address II: Developing Teen Minds in a Digital Age: Research and Innovation Toward a Future Science of Learning by Patricia K. Kuhl
Children watching videos of people talking don't respond with language development the way that they
do if a person is talking to them. They need social interaction.
Social experiences shape brain plasticity. MOOCs don't give you social interaction; teachers do.
Getting information from adult brains into child brains happens explicitly through talk,
but a lot of it is absorbed implicitly from living in a social culture. Humans have less innate information
than other species. Humans depend on others to learn. There are pros and cons to this. It allows depth of
friendship and love, but it is also makes us vulnerable to manipulation.
Sensitive periods in development are times with the brain expect experience from culture to
shape its circuitry. From 0-5, our brains are especially sensitive to linguistic information (this is true even
prenatally). They respond to the accent of the mother's accent differently than others. They'll attend to
languages other than the mothers because they are curious about language at birth.
The sensitive period for social interaction is 11-15. This will not be a shock to parents or teachers.
We now have more information about brain function, brain structure, and the interventions that enhance
malleable skills. People are not necessarily born "good at math" or "good at languages" or "good at music."
They have experiences that enhance their abilities in those areas.
Prior to six months, children of all cultures can make and hear the phonetic sounds of all languages. We
are born "citizens of the world." At 8 months, they get better at attending to the sounds that are happening
in the area in which they live. (Japanese children learn to ignore the distinction between r and l. American
children learn to ignore the distinctions in Mandarin.) Kids' brains are pruning away sounds they won't
need. At 10-12 months, they exposed American babies to Mandarin speakers. They quickly re-acquired
those sounds. They repeated it with equivalent videos, and it had no effect on their acquisition.
You get perfect learning in a complex live setting, and you get none in a video setting, even if they are
the same exposure.
When they put two babies in the room with the video, there was learning from the video. The social
interactions of the babies with each other caused them to learn. The biochemistry is different in a face to
face social interaction than when they are alone.Speech activates auditory and motor-speech areas.
Brain development in the prefrontal cortex (executive function) lags behind the limbic system (emotion), so the teen brain is led more by emotion. Kids take more risks with others than they do alone. Teen mental health issues often come from experiences influencing the limbic system.
March 2020 - Pandemic Lockdown
In the beginning, kids were valiantly interacting with the Zoom screens and were trying to learn well.
A few months in, the lack of social interaction and increased screen time really got to them.
Compared the thickness of the cortex of 9-17 year olds who had been tested in 2018 with subsequent
tests in 2021 (original plan had been to test in 2020, but the lab was closed). The increased the rate of
thinning over predicted models, meaning the brain aged faster than it should have (girls even more in all
lobes of both hemispheres while boys were only in the occipital lobe and at around half the rate). The
same effect held for white matter pathways. The rhythms showed accelerated aging in both genders.
The novel stressors of the pandemic have led to increased anxiety and depression. Some (especially those
with a structured routine, more time in nature, and adequate sleep) were more resilient.
What is hopeful about the future?
- We have better interventions for enhancing learning.
- Parents and teachers better understand the power of socialization in the brain.
- We can use this knowledge to create more flexible learners.
- Human ingenuity rises to the occasion in the face of difficulty.
The social brain is the gateway to cognition, and it drives learning!
Session: How to Ensure Students Not only Know What's True but Understand Why It's True by Daniel Willingham
It is easier to acquire knowledge than it is to develop understanding. We are even worse at assessing whether or not we understand something. But why is that so?
If you look for a definition of understanding, you are likely to find one that describes the consequences of understanding, not a definition for the word itself.
Most of the time, action is not driven by reasoning. There is an automatic association. (You don't reason, "It's raining. What's the best way to go about dealing with this. Perhaps, I could . . . ." No, you just get the umbrella.)
The difference between knowing and understanding, but there are also different levels of understanding. I think I understand my can opener (how the mechanism works) in a way I don't understand my hand vacuum (It's just magic as far as I'm concerned).
Views of Understanding
Interventionist view of causality - "If X changes, Y will happen." This is often the way we understand science, math, and history because there are responsive systems. What will happen to my can opener if I do . . .?
Contextual Understanding - Allowing a richer web of connections. A child's schema for whale includes fish because of all of their similarities. An adult comes along and tells them that a whale is a mammal, but does the child understand what it means to be a mammal? The more the child knows about mammals, the more they can infer about whales. They can ask things like, "Is there something else that I thought was a fish that might be a mammal?"
We tend to present information hierarchically - Main point, three conclusions, supporting evidence, details. We don't tend to structure as a web. We present in narratives because kids have a mental template for it already when they reach school. As things get more complicated, they start to be presented more hierarchically. But we expect kids to recognize the hierarchy even when some of the connecting things are separated in time.
Deep structure and surface structure can be hard to differentiate. Seeing past the details of the example (meat, bells, and salivation in Pavlov's dog) to the structure of classical conditioning (pairing 2 stimuli with an involuntary response) can be hard to tease out, especially for novice learners.
We are born equipped to understand some aspects of the world. We are also born with a bias to interpret them in certain ways. Much of what we learn in school is to show that some things go against our natural interpretations. (Just because something seems true does not mean that's how it actually works.)
How do we differentiate objects from their attributes? If a parent and child are standing in a field, and a rabbit hops by and the parents says, "Look a rabbit," how does a child interpret what the word rabbit means? This can be especially difficult when learning physics, which is why some of the medieval conclusions, like impetus, were wrong even though they kind of made sense. Children aren't blank slates; they come in with beliefs and biases that will affect what they get out of what you teach them.
Teleological understanding - With minimal stimulus, you think of cause, relationships, goals, and desires. It's what enables us to have theory of mind. This can lead to problems: We tend to see purpose where there is none. We tend to see agency where there is none.
We naturally understand numerosity, but not on a linear scale: We have understand of things one at a time up until about 5. Then, we start understanding them in groups of five. Then we use vague terms like few, several, some, heaps, and tons. We compare things on a more logarithmic scale where we understanding orders of magnitude more than individual components. (Example: The difference between 2 and 6 is huge - One is triple the other. The difference between 102 and 106 is minimal. In school, we then teach that the difference between them is the same (4), overcoming a natural bias toward logarithmic thinking.)
Manage your own expectations of student understandings - Since it is hard, figure out what are the core ideas that it is critical they understand and focus heavily on those, looping back around to them.
Help them make connections with your talk and with reading - Preview your organization at the beginning and point back to it during the instruction. Point to it and say, "I'm moving on to the second point here." If they are old enough that they are taking notes, encourage them to revisit and re-organize the notes from linear order to hierarchical order. (This will have to be modeled, or they will just think it is busy work.)
To see deep structure, get them to solve a problem and then generate a general solution that type of problem. Get them to solve a problem, then say, "What if we change this aspect? How will that change the solution?" Give two problems that are both about the same deep structure but with very different surface aspects and ask them what the problems have in common.
Good analogies are critical, especially with unnatural learning. Singapore math is great at this. They do great spatial modeling to build bridges across ideas.
How can kids know if they understand. Explicitly explain to them what it means to understand. It builds from passive to active to expansive.
- It means sense to make when it is explained.
- I can explain it myself. (Have them do a turn and talk when they think you understand but you think they don't.)
- I can apply it in new contexts.
Keynote Address III: The Science of Learning: Building a Culture of Cognitive Engagement and Learning by Doug Lemov
In The Curious Case of the Dog in the Night Time, Sherlock solves the case because a dog did not bark.
We will look at The Curious Case of the First Year Teacher Whose Lesson Went Almost Perfectly
Basketball video with the moonwalking bear. - It's easy to miss something you're not looking for.
Attention is fragile
- We are only aware of a amll portion of our visual world at any moment, but we are resistant to believing that. We insist that we are fully aware when we are not.
- Choosing to pay attention to one thing is inherently choosing to ignore others. Attention is selective.
- Attention is a pre-requisite to learning. You can't learn something that you don't pay attention to.
- Attention is contagious. It's a viral social phenomenon.
(Personal note: I get why Tracking works for attention, but it would weird me out as both a student and a teacher to have that level of sustained attention from a group. I think it would put me in fight or flight mode.)
"Learning can be defined as a change in long-term memory. If nothing has changed in long-term memory, nothing has been learned." - Paul Kirschner
Your intention to remember has little effect on your ability to do so. Your brain doesn't care what you want (I think Dan Willingham said that in the previous session). You can't just decide to remember something.
When your working memory is overloaded, your perception of the environment is degraded. This is why distracted driving is so dangerous.
Best ways to deal with working memory - Routines, explicit instruction, retrieval practice, fluency.
Teacher working memory is taxed quickly (especially first year teachers) - Embed routines, habits, and systems to help your own working memory. A visible and predictable routine is also a classroom management strategy because it becomes a social norm. It has the side effect of creating a sense of belonging. (Every faith sings or chants or recites together, and it makes them feel more connected to the other members of their faith.)
You can cue a routine with almost no load on your working memory or theirs.
No comments:
Post a Comment