Educator quality
I was recently asked to help lead an effort to better measure and improve the quality of education in my department. This post lists a few things I’ve been pondering about education as a consequence of that request. I’ve also been reflecting on the topic of my previous post Capturing the Wild Attention, but don’t have more to add on it at this time.
One part of an educator’s job is pacing.
You have a finite amount of working memory, which is used to hold all three of
Because working memory is finite, we rely on your subconscious providing high-enough-level ideas that you have spare room for learning with. Because you’re reading this post, I know you’ve already done this. You’re not having to use your conscious mind to figure out what letters the shapes on the screen represent, nor to puzzle out the words the letters form nor the meaning they convey when strung together into sentences. Instead, your subconscious presents you with high-level ideas as you read, freeing up your conscious mind to think in terms of those high-level ideas.
In well-paced learning, you spend enough time consciously constructing a new idea for it to become subconsciously constructed before moving on to the next ideas. If you try to learn at too fast a pace you experience cognitive overload: you need all your working memory to hold old ideas and none left over to construct new ones, meaning learning ceases. If you try to learn at too slow a pace you have no new ideas to construct and no learning takes place.
Different orderings of material allow well-paced learning at different speeds. For example, you could learn that 3 + 4
and 7
are different ways of writing the same idea before you learn what ideas 3
, +
, 4
, and 7
represent, but we’d expect learning arithmetic in that order would require more time and energy than learning the meaning of individual symbols first.
In self-guided instruction, the learner uses their own working memory to figure out what order to learn things in. In self-paced instruction, the learner has the order provided but uses their own working memory to figure out when they have learned enough to move on. In synchronous instruction, both of those are the job of an instructor, potentially freeing up more working memory to learn with but also increasing the potential for a mismatch between learner and instructor pacing.
The truth is too large to learn all at once. Hence, we have to learn something less than the truth first.
As one example, I took up etymology as a kind of hobby in my 30s. In doing so I learned that words I’d been using all my life like person
have intricate history and nuanced, context-dependent denotation and connotation. But I would never try to explain that nuance to a new learner of English: I’d tell them a simplified version of one of its common meanings and leave it at that. In other words, I’d deceive them, lie to them, tell them a falsehood—but one that is true
or at least true enough for them to learn more truth.
The challenge of figuring out truth-enough falsehoods that are more learnable than the full truth is ever-present in teaching. I’ve used dozens of hopefully true-enough falsehoods in this post so far: what working memory is, what it holds, how they interact, how you read, what good pacing is, how overload works, what happens when pacing is slow, what self-guided and self-paced instruction are, and so on. Did I pick the right analogies, the right balance between understandable and true? That is a question that never leaves an introspective educator.
In teaching computer programming, one class of these falsehoods is so important that it has a formal name: notional machine.
Real computers are engineered by humans to operate predictably and teaching all of how they work is possible, but it takes many months if not years of instruction to do so. So instead we tell students that the computer they are talking to works in a much simpler way, giving them an understandable notion of what the machine is doing that will help them learn programming. That notion is false, in many ways, but if we pick a good notional machine those falsehoods are positioned in places where they won’t discover them until they’ve been programming for months.
All teaching has something like a notional machine,
a simplified false proxy for the complex truth that is the context for learning. Picking the right falsehood, simple enough to learn from and true enough to learn useful things from, is an ongoing challenge.
Education is an exercise in delayed gratification. You spend hours or years filling your head with things you don’t care about or see a use for with the promise that eventually they’ll culminate in knowledge worth having. But if you stop believing in that promise, education becomes much more difficult. Nothing can ruin an educator’s ability to teach like losing their students’ trust.
Trust is also needed on a much smaller scale. When a student comes to my office for help, most often they are aware of being stuck on one topic but are stuck there because the pacing was too fast and there are actually other, earlier concepts they’ve failed to learn that they need to revisit before the topic that brought them to my office can be resolved. Hence, the first thing I need to do is get them to trust me that they can drop the thing they think is the problem and follow me in a discussion of something that seems to them to be unrelated.
Building trust is the aspect of teaching that I least know how to help others master. Most of the ideas I have about it are paired with an opposite idea I also have about it. For example, it helps to project confidence, but it also helps to admit you don’t know. Trust can be eroded by asking for trust, but it can also be strengthened by telling students to trust you. It helps to admit you’re picking the best pacing and structure you can but might have them wrong, but it hurts to seem uncertain about how to pace or structure things. And so on.
Trust is also something where some people have an innate advantage over others. Because we all learned to trust adults over children, and because adults tend to be taller, older-looking, and deeper-voiced than children, all of our subconsciousnesses will ascribe an unearned trustworthy
designation more to people who are taller, older-looking, and deeper-voiced. Your experience and the culture you were raised in may add additional trust biases, giving more unearned trusting to people who remind you of people you’ve trusted in the past.