Andrew began his classroom life as a high-school English teacher in 1988, and has been working in or near schools ever since. In 2008, Andrew began exploring the practical application of psychology and neuroscience in his classroom. In 2011, he earned his M. Ed. from the “Mind, Brain, Education” program at Harvard University. As President of “Translate the Brain,” Andrew now works with teachers, students, administrators, and parents to make learning easier and teaching more effective. He has presented at schools and workshops across the country; he also serves as an adviser to several organizations, including “The People’s Science.”
Andrew is the author of "Learning Begins: The Science of Working Memory and Attention for the Classroom Teacher."
Like so many who study psychology, we at LatB are terribly sad to learn that Walter Mischel has died.
The New York Times obituary describes his importance — both in revolutionizing the field of psychology, and in popular understanding of self-control.
Personality in Context
For psychologists, Mischel emphasized the importance of context.
Personality theory suggests that “I’m this kind of person, not that kind of person.” Mischel’s research emphasizes that “I’m this kind of person in these circumstances, and that kind of person in other circumstances.” Context always matters.
This insight can’t be over-emphasized among teachers who want to teach with research in mind.
We shouldn’t just hear about a particular psychology study and adopt its methodology. We should, instead, check to see if the method that worked in that context might also work in this context. Which is to say: our classroom.
After all: most psychology research happens with college students. If you don’t teach college students, you need to adapt that research to your context.
Marshmallows in Context
In popular culture, Mischel is best known for the “marshmallow test” (which, by the way, included other treats beyond marshmallows”).
You’ve seen videos of adorable 5-year-olds desperately trying to resist yummy goodness. For example:
https://www.youtube.com/watch?v=4L-n8Z7G0ic
The often-oversimplified test includes many nuances. For instance, the 5-year-olds respond differently based on how much they trust you.
But the headline remains important: self-control matters.
Those of us who got to see Walter Mischel speak at the November 2015 conference will not soon forget his clarity, thoughtfulness, and warmth.
He even helped Kelly Williams organize the very first Learning and the Brain conference.)
Dr. Fischer meant that even though we’ve been studying the brain for over 100 years, we still haven’t figured out very much about this infinitely complex part of our daily lives.
Years later, by the time I finished graduate school, he had started to admit we might be in Brain 1st Grade.
Exciting brain scanning techniques — especially fMRI — have shown us extraordinary and unexpected truths about our brains’ development and function.
Brain 2nd Grade?
Two recent discoveries make me wonder: are we in second grade yet?
First, a just-published study identifies a new brain cell — one that might, in fact, be unique to humans.
We don’t know much about “rosehip neurons”; for instance, we don’t know exactly what they do. They’re a kind of inhibitory neuron, but what they inhibit and why, we don’t yet know.
More than 100 years after Santiago Ramon y Cajal produced his amazing drawings of various brain cell types, we still haven’t identified them all.
Second, another just-published study suggests a new way for the brain to communicate with the body’s immune system.
The details here are quite complex (unless you’re already up to speed on aseptic meningitis and neutrophils).
But the surprise remains. Even now, we’re still figuring out basics: like, how does the brain talk with the immune system?
New Brain Cell: Classroom Implications
Of course, this news doesn’t yet tell us how to teach differently.
This new brain cell does, however, provide us an important reminder. Practically everything we learn about the brain in the 20-teens and 2020s will be tentative, initial, and incomplete. We should be excited with each development…and always ready to have old beliefs overturned by new findings.
Last week, I wrote about a potential strategy for making group-work more effective. A Boston-based research team has found reason to think that “intermittent” collaboration might yield better results than constant (or absent) collaboration.
Although I’m excited to see these results, my article concluded with two concerns:
First: the research was done with college students. It might not apply to younger learners.
Second: the participants weren’t exactly collaborating. They were (or were not) looking at each others’ answers after they solved problems.
They did not do what typically happens in schools, where students work on problems and projects together all at the same time.
So, again: this research might not (or, might) apply in our classrooms.
(I should be clear: the researchers don’t claim to be studying collaboration. Their research field is “collective intelligence.” The most obvious place to apply their research is in what teachers call “collaboration.”)
Thoughtful Update for Concern #1…
When I have questions about a study, I try to ask the researchers for their thoughts. In this case, I reached out to Dr. Jesse Shore with my two concerns.
His answers struck me as particularly helpful, and so I’m sharing them with you. (I’m rewording passages from his email, but with his permission drawing substantially on them.)
First, Dr. Shore explains why intermittent “collaboration” helps:
The results depend on (1) people trying more diverse solutions when they are notseeing others’ solutions [ACW: that’s how working alone helps]
and (2) learning from the solutions of others when they do see them [ACW: that’s how “collaborating” helps].
Intermittent collaboration makes time for both benefits.
For this reason, Dr. Shore suspects that this strategy would work well with younger students. There’s no obvious reason why 3rd graders (for example) wouldn’t “try more diverse solutions when not seeing others’ solutions,” nor why they wouldn’t “learn from others’ solutions when they see them.”
Such hypotheses need testing, but that’s a plausible set of presumptions with which to start.
…and Concern #2
Dr. Shore shares my concern about applying “collective intelligence” research directly to “collaboration.” After all, collaboration includes at least two other key variables.
The “free rider” problem. When groups work together, some people can just sit back and let others do all the work.
Interpersonal relationships. In groups, some people like each other and work well together; others just don’t. Or, a student might adopt another student’s strategy not because it works well, but because that student is popular.
Despite these other variables, Dr. Shore writes, “my guess is that intermittent interaction would still be best.”
After all–as I think about his summary–it seems clear that intermittent collaboration would interrupt the “free-rider” problem. I can’t let you do all the work if I have to work by myself at times.
The Big Picture
First: Dr. Shore offers us good reasons to think that “intermittent” interaction would indeed benefit typical kinds of school collaboration.
Its benefits, most likely, help students of all ages. And it might help with (and certainly wouldn’t exacerbate) the additional complexities of full-on collaboration.
Second: this strikes me as an excellent example of the philosophy that teachers shouldn’t just “do this thing,” but instead should “think this way.”
In this case: when we see research about “collaboration,” we should not simply enact its guidance. Instead, we should contemplate the specific ways it does, and does not, fit exactly with what we do.
Presumably, the study compares “the generation effect” to “rote memorization.” Presumably it concludes the former helps more than the latter.
Nope. Not a bit of it.
This study, instead, looks at different kinds of mistakes that students might make when they guess. It concludes that nearby guesses help, but far away guesses don’t.
Here’s the researchers’ summary: “errors benefit memory to the extent that they overlap semantically with targets.”
The study just isn’t about rote memorization. (Just to be sure, I checked with the study’s lead author. She confirms my understanding of the research.)
Lesson Learned
If you can’t even trust [name-of-well-known-education-blog], what can you do? How can you trust any news from the interwebs.
Here’s my advice:
First: anyone who quotes research should link to it. If you can’t find the link quite easily, don’t make any changes to your teaching.
Second: click the link. (Don’t yet make any changes to your teaching.)
Third: feel a little nervous. You’ve got several daunting pages of big words and bizarro graphs.
Fourth: Think about giving in and just making the changes that the article suggests. Don’t give in to that impulse. No changes yet.
Fifth: read the first paragraph of the research. It’s called the “abstract,” and it should summarize what the researchers did and what they concluded.
If the abstract DOESN’T include the point you read on the website, then you’re done. The research doesn’t focus on the argument that the blogger is making, and so shouldn’t have been cited in the first place.
Lesson learned. Nothing to see here. Move on.
In this case: when you read the abstract about the generation effect, you’d see that it never mentions rote memorization.
Of course, if the abstract DOES include that point — now you can start taking the blog seriously. You might not decide to change your teaching, but at least you’ve got an argument worth considering.
Flipping the Script
I had a similar experience (in reverse) about a month ago. This blog’s software notified me that another blog had linked to my article on omega-3 fish oil.
That post isn’t obviously the sexiest one on the website, so I was a bit surprised to see it getting internet love.
A brief investigation showed that the link came from a website in a foreign language — one that encouraged pregnant women to buy omega-3 fish oil supplements.
Hmmm.
My article summarized research showing that such supplements don’t help with working memory or in-school behavior. Not so much about benefits for pregnant women.
It seems clear that this other blog assumed its readers a) wouldn’t click on the link, and b) if they did, they wouldn’t be able to read English well.
The link was there to fool their readers, not help them.
The Headline
When we see a blog link to research, we can feel reassured. At the other end of the link, doubtless, lies research supporting the author’s point.
Does collaborative learning help students? If yes, what guidelines should teachers follow?
Collaborative Learning: Benefits and Detriments
Overall, we’ve got lots of research suggesting that collaboration helps students learn. And, happily, it doesn’t cost lots of extra dollars.
More specifically: the average score for students who learn in groups exceeds that of those who learn individually.
Unsurprisingly, students who struggle to learn benefit from practice with peers who understand better than they do.
At the same time, the highest scores tend to be lower in groups than among individual learners.
Working in groups, it seems, reduces the mental exploration necessary to find the best answers.
Given this background, we arrive at a really interesting question:
Can we get the benefits of group learning (higher average) AND the benefits of individual learning (highest scores).
It’s All in the Timing
Researchers at several Boston universities wondered if timing mattered. What would happen if students worked in groups at times and alone at other times?
The research team invited college students to work on a spatial puzzle. (It’s called the “Euclidean travelling salesperson problem.” I myself doubt that many of Euclid’s peers were travelling salespeople.)
Some of the students could always see their peers’ solutions. Some could never see those solutions. And some got to see every third solution.
Which groups progressed faster?
As they had hoped, the team found that the third group yielded both the highest average and the highest score.
In brief: teamwork helps most when team members also spend time working by themselves.
Classroom Implications for Collaborative Learning
This study offers a helpful suggestion. Teachers who use group work might ensure that group members work together at some times and solo at others.
At the same time, we should note some important caveats before we follow this guidance too strictly.
First: this study worked with college students. Its findings might apply to younger students. But, then again, they might not.
Second: this research is most easily described as “collaboration,” but that’s not exactly what the research team was studying. Notice: the participants never worked together on the travelling salesperson problem. Instead, they solved the problem on their own and then could (or could not) look at other students’ solutions.
That’s not typically how collaborative learning happens in schools.
More often, “collaborative learning” means that students work together on the project or problem. This study didn’t explore that approach.
(To be precise: the researchers focus on “collective intelligence,” not “collaborative learning.”)
Final Words
I myself think this research offers a helpful suggestion: occasional teamwork might lead to better results than constant (or absent) teamwork.
However, we should keep a sharp eye out for the actual results in our own classrooms. Unless you teach college students by having them look at each others’ correct answers, this study doesn’t explore your methodology precisely.
If you’re like most people, the answer is: “not very.”
We’ve got lots of research showing that people change their beliefs when they hear good news. However, they don’t change their views much when they hear bad news.
For example: I might ask you, “what are the odds that — in your lifetime — your house will be burgled?”
You answer “40%.”
Later on, I inform you that the real number is 30%. Given your initial estimate, I just gave you good news! You’re safer than you thought.
When I ask you the same question later, you’re likely to update your answer. You might guess 32%. That number is still high, but much more accurate than it was.
However, if you initially guessed “20%,” then the real number “30%” is bad news. You’re in more danger than you thought!
When I ask you the same question later, you probably won’t update your answer much. You’re likely to say “21%.”
To find out, they invited people to their lab and stressed out half of them.
(The stressed-out half heard they would have to give an impromptu speech in front of judges. And, they were given challenging math problems to solve.)
The researchers then asked them several questions like the one above: “how likely is it that your house will be burgled?”
How honestly did these participants process the correct information they got?
As before, the un-stressed participants learned from the good news, but not from the bad.
However, the research team found that stress helps. That is: the participants who worried about their upcoming public speaking gig processed the bad news as well as the good.
Next Steps
The research team double checked their results with fire fighters in Colorado. They got the same results. That’s helpful news.
However, all of this research focuses on adults. The average age in the first study was about 25 years. In the second study, 43 years.
We know that adolescents and children process emotions quite differently. So: we should cross our fingers and hope that the researchers try out their idea with school-aged children.
The more we understand the benefits as well as the detriments of stress, the better we can help our students navigate the appropriate challenges that school provides.
——————————————————————————————
For further thoughts on stress in schools, check out this earlier blog article by Rose Hendricks.
And, for fun, here’s a video of the lead researcher talking about some of his earlier work:
When I talk with my English students about The Glass Menagerie, we always identify the protagonist and the antagonist. This discussion helps them understand useful literary terms. It also clarifies their understanding of the play.
Of course, as they consider this question, I want them to recall a similar conversation we had about Macbeth. In that play as well, we can struggle to determine who the antagonist might be.
In psychology terminology, I want my students to “activate prior knowledge.” Their discussion of The Glass Menagerie will improve if they think about their prior knowledge of Macbeth.
Here’s the simplest teaching strategy in the world. If I want them to think about Macbeth‘s protagonist before they discuss TGM, I can start our class discussion with Shakespeare.
Rather than hope my students draw on their prior Macbeth knowledge, I can ensure that they do so.
This remarkably simple strategy has gotten recent research support. In this study, Dutch psychologists simply told students to recall prior learning before they undertook new learning. Those simple instructions boosted students’ scores.
Prior Knowledge: From Lesson Plan to Syllabus
This research advice might seem quite simple — even too simple. At the same time, I think it helps us understand less intuitive teaching advice.
You have probably heard about “the spacing effect.” When students spread practice out over time, they learn more than if they do all their practice at once.
To illustrate this idea, let’s look at a year-long plan in a blog by Mr. Benney:
As you can see, Mr. Benney teaches his first science topic in September. He then includes topic-1 problems in his students’ October homework (“lag homework”). He reintroduces the subject in December. And returns to it one final time in April.
Clearly, he has spaced out his students’ interactions with this topic.
But, notice what happens when he does this with all eight topics:
For many teachers, May looks quite scary indeed. Students are learning topic 8. They’re doing lag homework on topic 7. They’re being reintroduced to topics five and six. And they’re being re-re-introduced to topics 2 and 3.
Six topics all at the same time?
And yet, spacing requires interleaving. If Mr. Benney spreads out topic 1, then it will automatically interleave with the topics he’s teaching in October, December, and April. You can’t do one without the other.
Why would this be? After all, May’s syllabus looks really complicated.
Perhaps recent research on “prior knowledge” explains this result. If students are thinking about several topics at the same time, then their prior knowledge from previous months remains active.
Macbeth isn’t something we talked about 3 months ago. We have talked about it several times, including just last week.
Here’s the equation. Spacing automatically leads to interleaving. And, interleaving in turn keeps prior knowledge active. These three teaching strategies combine in multiple ways to help our students learn.
As the school year begins, we all want our students to learn more stuff.
We want them to learn phonics rules, or multiplication tables, or Boyle’s law, or the importance of the 13th amendment.
We also might want them to learn more general skills.
We’d like them to learn how to learn. Or, how to manage their emotions. Or, how to focus on one thing at a time.
Must we accomplish our goals by teaching each of those topics specifically? Or, can we teach students one basic skill to help them learn everything else?
“Brain Fitness”? Cognitive Training?
For example: almost all athletes need to have a strong core and high levels of aerobic fitness. If, as a coach, I focus my work there, I help my players get better in almost all sports simultaneously.
Is there a brain analogue for “core strength and aerobic fitness”?
We’ve got lots of research showing that chess players score better on IQ and working memory tests than the general population.
Does it follow then that chess training increases general intelligence? If yes, then chess lessons would help students learn to read, and solve quadratic equations, and understand mitosis.
The research giveth…
In the short term, the answer is: “just maybe yes!”
These results may be considered “cautiously promising.” In fact, they are not. The size of the effects was inversely related to the quality of the experimental design. Specifically, when the experimental groups were compared with active control groups — … to rule out possible placebo effects… [or] the excitement induced by a novel activity — the overall effect sizes were minimal or null.
In other words: the better the research, the less likely it was to show any benefit. Almost certainly, general cognitive training led to improvement only because participants believed it would.
Practical Implications
The bad news: we just don’t have good evidence that chess, or working memory training, or music lessons improve other cognitive abilities.
(Of course, chess lessons make people better at chess. Oboe lessons make people better at playing the oboe.)
The good news: school works. When we want our students to learn how to analyze a poem, we can teach them to do so: one beautiful poem at a time.
Psychology can offer advice to teachers, but sometimes that advice is frustratingly vague.
We know, for example, that attention is important.
We know that it results from a combination of three neural processes: alertness, orienting, and executive attention.
But: what do teachers DO with that knowledge? How do we teach any differently?
Retrieval Practice Timing
For example, we’ve seen lots of research showing that retrieval practicehelps students learn.
That is: rather than simply looking back over material, students should somehow quiz themselves on it.
They might use flash cards.
They might try a “brain dump”: just writing down everything they remember on a topic.
They can use quizlet to review key points.
Given that lots of student learning happens with textbooks, teachers have a very practical question: when should they do that retrieval?
Textbooks often include practice questions. Should students try to answer them as they go along? Or, should they wait until they have read the full chapter?
Happily, this question can be studied quite straightforwardly.
A third group answered the self-study questions in the textbook. Some of those questions appeared at the end of a section. Some appeared at the end of the chapter. And some appeared in both places.
Which group remembered this information better two days later?
Retrieval Practice Timing, and Beyond
Unsurprisingly, the students who reread the information remembered a bit more than those who did not. The rereaders averaged a 44.8 on the quiz, whereas the one-time readers averaged a 34.2.
The self-testers? They averaged a 61.5.
As we’ve seen before, self-testing is a HUGE help.
(By the way, you might think “61.5 is a terrible score. That’s almost failing!” However, this isn’t a class test; it’s a relative measurement. The point isn’t what the students scored, but how the groups scored compared to each other. The self-testers remembered much more.)
What about the timing? Is it better to answer questions at the end of the section, or the end of the chapter.
As it turned out, both times worked equally well. As long as students do the retrieval practice, it doesn’t particularly matter at what point in the chapter they do so.
Here’s the intriguing finding: questions answered twice — both at the end of the section and and the end of the chapter — led to even higher learning.
Psychology researchers could easily get focused on studies in the lab. They can control variables better; they’re faster to run.
I always feel especially happy to find researchers who keep their gaze on practical classroom applications.
In this case, we’ve learned: a) that retrieval practice helps students learn from textbooks, b) that students can answer relevant questions at any time and still get this benefit, and c) that two attempts to answer the question are (or, at least, might be) better than one.
About Andrew Watson 
