Author Archives: Andrew Watson

Here’s a common classroom problem.

As I’m explaning a complex concept, a student raises a hand.

“Just a moment,” I say, and finish my explanation.

Now I turn and smile at the student: “what was your question?” I ask.

All too often, the student answers, “I forgot my question.”

What’s going on here?

As is so often the case, the answer is: working memory overload.

Working memory HOLDS and PROCESSES information. When a student fails to hold and process, that’s working memory overload.

In this case, my student was processing my explanation, and so failed to hold the question.

The solution?

It might seem simple. Don’t ask students to hold questions while they process explanations.

Instead, I should answer students’ questions right away. Problem solved….

When Solutions Create Problems

Wait just a moment.

This “solution” I just offered might solve the student’s problem.

At the same time, it might create new problems.

The student’s question — even a well-intentioned one — might throw my explanation off track.

My students might lose their tentative understanding of my complex explanation.

I might lose my own train of thought.

So I fixed one classroom problem but now have yet another one. YIKES.

What’s a teacher to do?

First Things First

This example — but one of many — might throw our entire project into question.

Teachers turn to psychology and neuroscience to solve classroom problems.

However, if these “research-based solutions” simply transform one problem into some other headache, why bother with the research?

We could save time by sticking with the old problem, right?

I think the fair answer to that question is: “actually, no.” Here’s why…

Teachers don’t need research to solve classroom problems. We need research to solve COMPLEX classroom problems.

When our classroom problems are simple, we just solve them on our own. We are — after all — teachers! We specialize in problem solving.

For that reason, we turn to research only when the problem isn’t simple.

And for that reason, we shouldn’t be surprised when the answer isn’t simple either.

OF COURSE we can’t fix the “questions-interrupting-my-explanation” problem with one easy research-based step.

If it were so simple a problem, we would have solved it without the research.

Changing the Lens

As I’ve explored this question with wise teachers in recent weeks, I’ve been struck by a pattern:

PROBLEM ONE requires SOLUTION ONE.

But: SOLUTION ONE creates PROBLEM TWO.

And: it’s often true that PROBLEM TWO comes from a different cognitive function than PROBLEM ONE.

So, in the example above, I started with a working memory problem: my student coudn’t hold and process information.

My solution (“take questions right away”) created another problem — but not a working memory problem.

When I answer questions mid-explanation, my students lose focus. That is, the working memory problem has been transformed into an attention problem.

To solve this second problem, I need to switch from working memory solutions to attention solutions.

In other words, I need to think about a separate cognitive function. I’ll find solutions to this 2nd order problem in a different research field.

Again with the Mantra

If you’ve ever heard me speak at a Learning and the Brain conference, you know my mantra: “don’t just do this thing; instead, think this way.”

In other words: psychology research can’t provide teachers with a list of “best practices.” The strategy that works in my 10th grade English classroom at a boarding school might not help 1st graders add numbers in a Montessori program.

But: the thought process I follow with my 10th graders might lead to beneficial solutions for those 1st graders. The answer (“do this thing”) might be different, but the mental pathway (“think this way”) will be the same.

The point I’m making here is: these thought processes might require us to leap from mental function to mental function in search of a more successful solution.

A solution to a long-term memory problem might uncover a motivational problem.

The solution to an alertness problem might promt an orienting problem.

When I reduce my students’ stress, I might ramp up their working memory difficulties.

And so forth.

When we understand research into all these topics, we can anticipate that these solutions might unveil an entirely different set of troubles.

And by moving nimbly from research topic to research topic, we can ultimately solve that complex problem that once seemed intractable.

All this nimbling about takes practice. And, ironically, it might threaten our own working memory capacity.

But once we get used to thinking this new way, we will arrive at solutions that fit our classrooms, and that work.

Consider the following paradox:

Teachers need to give students instructions — of course we do!

After all, instructions help students do what they need to do, so that they can learn what we want them to learn.

At the same time, too many instructions might very well overwhelm working memory.

After all, the student has to HOLD the instructions in memory while PROCESSING each one individually. And: “holding while processing” is one handy definition of working memory function.

In brief: the right number of instructions can help learning, but too many instructions can impede learning.

I recently asked a group of wise and experienced teachers this question:

“Can you think of other teaching practices — like instructions — that are beneficial in small amounts, but might create working memory overload in large amounts?”

After some time to think and discuss, one teacher answered: group work.

After all, he mused, collaboration might simplify some mental processes. But collaboration itself creates additional mental taxes — all that negotiating and delegating and coordinating and explaining.

And disagreeing.

And resolving.

Are they ways that teachers can reduce those “mental taxes” so that students get the benifits without the penalties?

If only we had a research-based answer to those questions…

Inspired by this teacher’s observation, I hunted up this study.

Quadratics in Quito

To explore this question, researchers working in Quito, Ecuador worked with 15-year-olds solving quadratic equations.

Specifically, they wanted to know if practice collaborating helps students collaborate effectively.

As is always true, research design gets tricky. But the overall design makes sense.

Some students did practice solving quadratic equations collaboratively; others didn’t.

For a second round of math learning, all students were then sorted into groups for collaborative learning.

So, did students who practiced collaborating do better on later collaboration?

For complex equations: YES. Both three days later and seven days later, students who practiced collaborating did better solving problems than students who didn’t.

For simple equations: NO. If the mental work wasn’t very hard, students didn’t need to practice to collaborate effectively.

In light of these findings, the researchers’ recommendations make good sense.

If learners are novices, learning tasks are complex, and information distribution demands [extensive cooperation], teachers should prepare group members … using similar problems.

…

If task information does not demand [extensive cooperation], it is not necessary for the teachers to prepare learners to collaborate.

I want to highlight one part of this summary: “using similar problems.”

This research team emphasizes that “collaboration” is NOT a general skill. Collaboration will look different depending on the precise demands of the discipline and the topic.

So: students who “practiced” were given a VERY specific format for learning how to collaborate on this task.

If we want to get the benefits of practice for our own students, we should be sure to tailor the practice in very specific ways.

The Story Behind the Story: an Analogy and a Principle

A research article like this study always begins with a summary of earlier research findings, conclusions, and questions.

This summary includes a particularly helpful foundational inquiry.

When does collaboration increase WM load so much as to threaten learning?

When does collaboration reduce WM load enough to promote learning?

Is there some sort of taxonomy to consider or principle to explore?

To explain, I’ll start with an analogy.

Imagine I want to illuminate a yard at night.

For all sorts of reasons, it would be simplest to have one lamp to do so. Having multiple lamps just adds to the complexity and expense of the project.

So, if my yard is small enough to be covered by one lamp, then I should use one. Adding more lamps makes the project worse — more complicated, more expensive — not better.

But at some point, a yard gets big enough to need multiple lamps. If I use only one lamp, I just can’t illuminate the full yard.

In this case, the additional expense and complexity of having multiple lamps provides a meaningful benefit.

You can see where this is going.

Here’s a potential principle:

If a cognitive problem is simple enough, then one student can solve it on her own.

Adding other students (“collaborative leaning”!) increases the WM complexity of the situation without providing any additional benefit.

In this case, each student’s mental effort has become less effective, not more effective.

If, on the other hand, a cognitive problem gets complex enough, then it goes beyond any one student’s cognitive capacity.

In that case, the problem benefits from additional students’ cognitive efforts — even though all those extra students do increase the complexity of the problem.

At some tipping point, when a problem gets complicated enough, it needs to be divided into sub-tasks — despite the complexity of managing that work.

At that tipping point, well-structured and precisely-practiced collaboration probably is more beneficial than harmful.

TL;DR

Groupwork (probably) increases WM demands on simple cognitive tasks, but reduces WM demands for complex cognitive tasks.

To get the most benefits from collaboration, students should practice that skill — and teachers should tailor the practice to the precise demands of the cognitive work.

---

Zambrano, J., Kirschner, F., Sweller, J., & Kirschner, P. A. (2019). Effects of group experience and information distribution on collaborative learning. Instructional Science, 47, 531-550.

Let’s face it: teaching is hard.

I’ve been a classroom teacher for roughly 20 years — how do I count summer school? — and I still find the work exhillarating, exhausting, baffling, uplifting, frustrating, humbling, and joyous.

And that was Tuesday.

And: I think I’m not the only one who finds teaching to be an extra-ordinary challenge. I mean, don’t get me wrong, I love it. But GOSH it’s hard.

This hard-won experience leads me — and perhaps you — to two conclusions:

First: people who haven’t taught in the classroom don’t fully understand the challenges of the work.

Until you’ve tried to follow a scrupulously-devised lesson plan despite the fact that an un-announced fire-drill is in progress, two students have switched sections, three don’t have their notebooks, and four don’t think the cell-phone policy applies at just this moment…you just don’t really know.

How could you? It’s “Misson Impossible: Chalkdust” in here.

Second: I need all the help I can get.

No, really.

You’ve got some research that might …

… help me create a more effective lesson plan?

… explain how attention really works?

… suggest study strategies to help my students learn?

… foster motivation, at the beginning of a lesson on grammar?

I’m all ears. Please. I’m practically begging here…

My Learning and the Brain Journey

I attended my first LatB conference in 2008; the topic was The Science of Attention.

I IMMEDIATELY realized that this conference was just what I needed. So much wisdom and advice. So many compelling suggestions.

And, so many graphs and pictures of brains!

I returned to the classroom and started rethinking everything.

What should be rewrite policy be?

Should my classroom decorations be in primary colors?

What’s the right number of new vocabulary words to teach per class?

I had research-y answers to all those questions.

After several years of attending conferences, I went back to grad school and got a degree combining education, psychology, and neuroscience.

And, in addition to teaching, I started training other teachers.

Now I offered LOTS of advice of my own:

Because working memory does this, teachers should do that.

Because long-term memory benefits from this, teachers should do that.

Because stress affects this…you get the picture.

Many teachers appreciated all this guidance. But some obviously didn’t.

For whatever reason, they just didn’t want to do what I was telling them to do!

I was genuinely surprised; after all, I knew I was right because the research said so!

“The Big Ask”

Over time, I’ve come to realize that those teachers didn’t want to do what I was telling them — and “what the research said” — because I had forgotten the first lesson described above.

I mean: yes (lesson #2), “I need all the help I can get.” So, research-based advice MIGHT help.

But also,

Yes (lesson #1), “people who haven’t taught in the classroom don’t fully understand the challenges of the work.” So, research-based advice might not apply to this classroom, this topic, this teacher, or this student.

In other words,

When I say to teachers, “you should change the way you teach because research says so, ” I’m making a REALLY BIG ASK.

After all, those students are ultimately their responsibility, not mine.

Yes, I have taught — but I teach English to 10th graders at a very selective school. I should be very careful when offering guidance to, say …

… 2nd grade teachers who work with struggling readers, or

… those with lots of students on the Autism spectrum, or

… teachers who work in different cultures.

Because I do know research well, and I know classrooms fairly well, I can make those Big Asks. But I should be humble about doing so. And I should be respectful when a teacher says, “your ‘research-based advice’ …

… conflicts with our school’s mission, or

… might work with college students, but probably won’t work with 2nd graders, or

… requires personality traits that I don’t have.”

Yes, I think teachers should listen thoughtfully to the guidance that comes from research. And, I think those of us who cite research should listen thoughtfully to the classroom specifics, and the experience, of teachers.

Teachers who resist ‘research-based advice’ might seem “stubborn,” but they also might be right.

The Story Behind the Story

In last week’s blog post, I summarized research about using gestures to teach specific science concepts. I also sounded a few notes of caution:

I don’t fully understand the concept of “embodied cognition,” and

I worry that a few very specific studies will be used to insist that teachers make broad changes to our classroom work.

Even as I wrote that post, I could hear colleagues’ voices in my head: “why are you always so grouchy? Why don’t you listen when people with PhDs say ‘this is the Next Important Thing’?”

The answer to those question is this week’s blog post.

I’m ‘grouchy’ because I worry our field is constantly making Big Asks of teachers.

We often make those Asks without acknowledging a) the limits of our research knowledge, and b) the breadth of teachers’ experience.

I am indeed optimistic about combining cognitive psychology research with teacherly experience to improve teaching and foster learning.

To make that combination work, we should respect “stubborn” teachers by making Respectful Asks.

Teachers, of course, ask students questions. ALL THE TIME with the questions.

We ask questions DURING a lesson in order to “check for understanding.”

We encourage students to ask themselves questions AFTER class, because “retrieval practice” promotes learning.

And, we ask questions BEFORE a unit — for at least two very good reasons.

In the first place, we need to know what our students already know. If we don’t evaluate their prior knowledge, we struggle to build on that prior knowledge in a coherent way.

In the second place, we have increasingly strong research about the benefits of “prequestions.”

Unlike “checks for understanding” and “retrieval practice,” “prequestions” come before the unit.

And unlike “measuring prior knowledge,” “prequestions” deliberately focus on facts and procedures that students don’t yet know.

So: if I’m teaching a unit on Their Eyes Were Watching God, I might ask my students:

“What is the definition of a ‘bildungsroman’?”

“Describe the friendship between Langston Hughes and Countee Cullen.”

“What does hair often symbolize in literature?”

Truth to tell, it’s quite unlikely that my 10th grade students know the answers to these questions. So: those are prequestions — not checks for understanding, or retrival practice, or confirmations of prior knowledge.

Here’s the headline: we have reason to believe that “prequestions” — used correctly — help students learn information.

Here’s the story…

Hot Off the Presses

Long-time readers know that Dr. Elizabeth Ligon Bjork has done LOTS of essential work in the field of long-term memory formation and “desireable difficulties.”

And, you know my admiration of Dr. Nick Soderstrom, whose distinction between “short-term performance” and “long-term learning” should inform all teachers’ discussions.

So: when the two work together, they have my attention!

In this case, they JUST published a study on the topic of “prequestions.”

And, this study took place in actual college classrooms — not simply in a psychology lab. For that reason, its conclusions have a better chance of applying to the real-world work that other teachers do in classrooms.

In this research, students answered prequestions at the beginning of a few lectures. The subsequent lectures then provided answers to those questions. (By the way: students got only about 1/3 of those prequestions right — so for the most part they didn’t know the answers.)

On the final exam, students had to answer questions that …

… DIRECTLY related to those prequestions, or

… INDIRECTLY related to those prequestions, or

… were NOT related to the prequestions.

Sure enough, they did better on both directly and indirectly related questions, compared to the unrelated questions.

In brief: prequestions really did help college students learn in the classroom.

So simple! So effective!

So, Those “Details”?

My title promises that we need to “get the details just right.” In this case, as in so many others, I have thoughts. (Important note: at this point, I’m switching from reporting on research to offering my experience-based opinions.)

First Thought

Soderstrom and Bjork specifically write that prequestions helped because students took them seriously.

Here’s my concern: while college students may have the metacognitive perspective to take prequestions seriously, I do worry that younger students might not.

That is: once younger students realize that their answers to these questions don’t really matter, they might not take them as seriously as their college-age selves would.

The structure of prequestions, in fact, might discourage seriousness. Students rarely know the answers to these questions — that’s the point. Why would students attend seriously to questions they can’t possibly answer?

This potential problem leads to two tentative suggestions:

TELL students how and why prequestions might help, and

Use prequestions only RARELY.

After all, the more often that students must answer un-answerable questions, the less likely they are to give them appropriate mental effort.

My hope is: students who encounter prequestions only rarely won’t get cynical about trying to answer them.

Second Thought

If we use prequestions only rarely, are some times better than others?

My instincts are: yes.

Simply put: use prequestions at the beginning of a unit to highlight the most important concepts.

If we can get the benefit of this technique only rarely, then use it at the most important times.

This advice comes from common sense, not  from research — but common sense isn’t entirely forbidden on this blog.

Third Thought

Not all prequestions are created equal.

If a prequestion forces a student to think — that’s a good prequestion: even if they get a wrong answer.

However, if a prequestion activates a prior misconception, that question will actively interfere with learning.

For that reason, we should follow this rule:

Ask prequestions where students don’t know what the answer is, and where they don’t wrongly believe that they do know what the answer is.

For instance:

If I ask my student “which falls faster: a 10-pound bowling ball or a 15-pound bowling ball,” they almost certainly …

… don’t know the correct answer (that’s good), but

… wrongly think that they DO know the correct answer (that’s bad).

So: that prequestion would activate a prior misconception — and make learning harder.

On the other hand, those prequestions I asked at the top of this post (definition of “bildungsroman”) almost certainly don’t active prior misconceptions.

A Secret Unveiled; A Plea for Teamwork

I confess I have one deep frustration with this research pool.

Almost all teachers — and all students — hate tests.

So: if I name something “the testing effect,” teachers and students will HATE it — even if it’s beneficial. (Hint: the “testing effect” is just another way of talking about “retrieval practice.”)

And, if I name something “pretesting,” teachers and students will HATE it — even if it’s beneficial. Pretesting sounds like a test, no?

Sure enough, researchers have named a beneficial teaching “pretesting,” thereby ensuring confusion, and discouraging its use.

But — of course — “pretesting” simply means “asking questions on a topic before you’ve taught the material.” It’s NOT A TEST. It’s just a set of QUESTIONS.

So, I’ve been writing about “prequestions,” although everyone else in this field calls them “pretests.”

I hope you’ll join me in this virtuous rebranding.

TL;DR

Prequestions (aka “pretesting”) help students learn new material — and not just the information in the questions themselves.

Because the technique works if students take it seriously, I suggest …

… using it rarely,

… using it for important material, and

… asking prequestions that DON’T activate prior misconceptions.

---

Soderstrom, N. C., & Bjork, E. L. (2023). Pretesting Enhances Learning in the Classroom. Educational Psychology Review, 35(3), 88.