Student-Led Classrooms Waste Teacher Skill

A slew of research shows that direct instruction produces superior results compared to other instructional methods. A new study in the Journal of Labor Economics by Eric Taylor provides more information on how and why. Using a randomized controlled trial, Taylor compares a weak form of direct instruction with student led classrooms in which:

the students are expected to reason through and articulate math concepts with each other,while teachers facilitate conversationsand help students express their thoughtswith a focus on [students] understanding, rather than on students answering problems correctly

He finds that direct instruction results in greater student learning. More importantly, however, he also has data on how well teachers understand math and how to teach math and what he finds is that this knowledge is basically only productive when teachers use direct instruction. In other words, teacher skill only produces results when teachers are assigned a task that uses that skill. Student-led classrooms waste teacher skill and so are less productive.

Hat tip: Jose the (Not) Mediocre.

Comments

'Student-Led Classrooms Waste Teacher Skill'

And this is a surprise why?

The Montessori method followers might dispute that, though they tend to have a different idea of teaching.

At least in preschool, Montessori classrooms are not student-led, although the students may be given some illusions of autonomy.

Yes, true. Montessori is a number of things that are still considered to fit under a single label, and most of my tangential experience of such is in Germany, where the overlap between Montessori and Waldorf/Steiner definitely exists, plus the whole (now far out of favor) aspect of anti-authoritarianism, ends up leading to a different perspective.

Student led is a bit ambiguous too - to my understanding, if a student in a Montessori framework is interested in something, that interest is followed - that is, the student(s) are the ones leading the curriculum, if not the classroom.

However, the illusion of autonomy is much less in Germany, but then I would need to start talking about Waldkindergartens in that regard - https://en.wikipedia.org/wiki/Forest_kindergarten Along with the reality that no one leaves a 4 year old completely unattended for hours at a stretch, regardless of what the 4 year old might believe.

Yes, as someone who has experienced such methods I'm not at all surprised.

There is a misunderstanding w/r/t Montessori and the term "student-led." It is better called "student-centered." And it is about meeting the kids where they are, and allowing them a say in how they proceed.

It is a clear juxtaposition to the caricature of mainstream public education's lowest-common-denominator one-size-fits-all image.

Hidden within that though, are in fact pretty explicit sequential instructions on how to break down lessons into digestible steps.

Done properly, it involves observant teachers circulating in small classrooms and making sure the plates are all still spinning.

I suppose if done poorly, it can resemble anarchy.

Yes. My understanding of Montessori is that it is the student deciding what to focus on, and the instructor then provides direct instruction in that thing. Not students teaching other students.

How about kids deciding to smoke pot and play chess?

Ask the local middle school about vaping....

Two kids who have been through Montessori, precisely because they are much more directed than public school. The activities are very step by step and in our very good Montessori the teachers are well-drilled.
But if the student wants to do math activities for a couple of months, thats fine. Its directed but not regimented.
Tyler needs to read the educational realist critique of DI. Its devastating.
https://educationrealist.wordpress.com/2018/03/28/four-obvious-objections-to-direct-instruction/

There is a big difference between direct instruction (lower case) meaning that the teacher directly instructs the students and Direct Instruction (capital letters), a scripted step-by-step program developed by Siegfried (Zig) Engelmanm and colleagues. The Education Realist post is about capital letter Direct Instruction. I'm pretty sure this post is about generic direct instruction.

Nope. The link that Alex provided in the description is to a previous post on the benefits of Direct Instruction developed by Engelmann.

Not surprised at all.

Student-led classrooms need motivated students. I remember 2 or 3 classes when this happened.

The rest of time there was a class clown, the guy who just wanted the HS diploma, and later the varsity team airhead. Does a knowledgeadble teacher really thinks one can "reason through and articulate math concepts" with any of these individuals?

This system might work in environments where students are thoroughly filtered, but not for most classrooms in the real world.

What even is this student centered Math instruction. I always remember teachers walking us through proofs. Understanding and reproducing a proof seems like it requires students to "reason through and articulate math concepts". Are high school students today all supposed to independently invent calculus or something?

I believe that is what my college level calculus class thought. Everything was student led. I dropped that class within a week and got into one with a fantastic teacher. Learned the concepts within minutes each class.

When the 7th grade algebra teacher's presentation to parents at open house was suggesting something like that, I asked for clarification, "Are the students supposed to talk together and come up with how to factor quadratic polynomials?" When it was affirmed that that was the plan, I thought of the centuries of multinational activity that led to that development. It seemed doubtful my child's class would get far on their own.

Funny, in mathematics it is known as the Moore method. It produced some of the best mathematicians of the 20th century.

You mean pedagogic fads which allow school employees to slack-off are not effective? Say it ain't so...

It's less about allowing teachers to slack off, and more about schools being able to blame their past mediocrity on using the wrong pedagogy (read:fad).

Student-centered instruction is no-way-no-how less work. You have to write the tasks to the level of your students, and you have to instruct small groups socratically while monitoring others. At the extreme, you have to develop different plans for different groups of students who are at different levels to follow simultaneously (my district expects this). Direct instruction is much easier in terms of the management, it's easier to see who is paying attention / on task and who isn't from the front of the room.

"...Teacher skill only produces results when teachers are assigned a task that uses that skill."

This might explain why ed-school graduates are so enthusiastic about student-led classrooms, and so averse to direct instruction: the latter would cruelly expose just how many people with education degrees know little or nothing about the subject that they're teaching.

"pedagogic fads"

exactly correct

'Pedagogue - Led Classrooms Waste Student Lives'

But it sounds like the difference between direct and indirect isn’t so great at the lower end of the teach math distribution. So maybe for bad teachers indirect instruction still makes sense.

I think it was that under "student-led" all outcomes were equally bad, controlling for teacher quality, and given a bad teacher, students would be better off under direct instruction than student-led.

My magnet school was based on a 20 min lecture followed by 40 min of the teacher going from small table (about 4 kids) working through problems together and offering assistance. This was vastly superior to 60 straight minutes of lecture. Maybe it just works for high school kids with high IQ.

Maybe. What did you do at the small tables? Math problems? That's probably productive assuming you couldn't or wouldn't have found a lower opportunity cost time to do them at home.

The worst is when they apply this stuff to history. Obviously the best way to learn history is to read a book or listen to a lecture. Experts still argue radically different interpretations of the fall of the Western Roman Empire. Some kid is not going to figure out what happened and why by pretending to be a slave having dinner with Theodoric.

"Experts still argue radically different interpretations of the fall of the Western Roman Empire".
How many of them are covered by the typical high school lecture? Why not just having students read their books and do exercises instead of having a teacher who probably know little about Roman History himself drone on and on about it? That could have made sense when there were no print presses or internet, when media was scarce and you had to hire a highly learned person (i.e someone who had been in the same room books before) to lecture you on things. Why not just schools to grade students and make sure they are gathered where their behavior and study habits can be observed and corrected? Why do we need in the dawn of the 21th Century human cassette tapes?

Yep. Nobody would actually track down a high school teacher if they want to learn something. There are some standouts, but most have pretty limited content knowledge. Worse, many don't know what they don't know. The only issue with kids just watching online lectures from Yale Profs and reading Bryan Ward-Perkins is that most kids wont do it, and to the extent that we care about low-to-average-IQ kids knowing these things, they would need it significantly dumbed down.

"Profs and reading Bryan Ward-Perkins is that most kids wont do it."
Most kids are not paying that much attention to their teachers anyway. If atrendance is the issue, we certainly can herd students to seat before computers or big screens or before books as well as we herd them to classes right now.

"to the extent that we care about low-to-average-IQ kids knowing these things, they would need it significantly dumbed down."
To be fair, they already are not learning most of those things, so we do not need much effort to get net positive outcomes. Maybe we can use all the money and manpower we thrown into highschool lectures to track children's understandment and concentrate effort at those who are not learning (the other can basically take care of themselves).

"This was vastly superior to 60 straight minutes of lecture": in my years at school I'll bet I was never once subjected to a sixty minute lecture. It would be a mad way to attempt to teach schoolchildren. I've never been convinced it's a good way to teach undergraduates. Or postgraduates.

(I may be biased: by far the worst teaching I ever encountered was in a postgraduate course. The chap was so bad that I continued to attend partly because I wanted to see if he could get any worse.)

"Nobody would actually track down a high school teacher if they want to learn something." I would: I can think of half-a-dozen of mine who were very good to excellent - knowledgeable, good explainers, and capable of getting us to discuss our work at something approaching an adult level. I admit I'd run a mile before I took instruction in Latin or Physics from the fellows who attempted to teach me those. Happily you can learn a lot of physics from books.

There is a contradiction here. The research purports to have "data on how well teachers understand math and how to teach math." But it then purports to find that one way most people in the ed business think is a good way to teach math is actually a bad way.

You can't have it both ways.

From the abstract: 'Teacher productivity—measured with student test scores—is increasing in math skills when teachers use conventional “direct instruction”'. From the paper: "whether and how a teacher’s math skills contribute to her productivity depends on how she is asked to teach math." The paper has data on how well the teacher understands math and data on how the teacher is told how teach (in this case, direct instruction vs. student led). In your mind, the phrase "how to teach math" is translated as "how well to teach math". No contradiction.

What is the contradiction, exactly? There was a time when most of the things medical professionals were taught to do were bad for their patients health. That is no stranger than discovering that most of the things education professionals are taught to do are bad for their students education.

The intelligence of a group of children is the inverse of it's sum.

Unlike adults....

That is why design by committee would so well.

Result depends on how "learning more" is assessed. If the instruction method focuses on answering problems correctly, and the test/assessment focuses on the same, then no surprise. If understanding (as opposed to problem-solving) were tested, the result might be different. That said, there is a place for both approaches, although my style is certainly much closer to direct instruction...

I don't have a link handy but the DI literature does cover new and novel problems for the students. DI still performs better.

"If understanding (as opposed to problem-solving) were tested, the result might be different." Given that test scores were lower in the student-led treatment, I don't think there was too much "understanding" going on.

I taught physics and tried to test/assess understanding. It was very difficult and took an astounding amount of grading time.

Think about it: How would you test/assess understanding? How would you do it for 100 young people?

That's why easily scored tests are so popular with the people who have to grade them. As the law professor saying goes, "I teach for free. They pay me to do grading."

It might be useful to do a similar study on the case method, where students discuss and argue about cases, with the instructor merely posing questions. Is this less effective that direct instruction?

One thing to note on the case method is that, in many first-year classes (whether case method is most prominent), the point is not so much to teach a student the black letter law, but rather to teach the student legal argument and how to analyze a case to determine what it says about what the law is. In practice, and in particular with respect to the common law, nobody actually uses law learned in law school, but rather they have to argue for legal conclusions based on analyzing cases.

Thanks.

I was thinking about the "Paper Chase" law school method where the teacher is very much the Sage on the Stage but much of the speaking in the class is done by students who are rather scared of the teacher. Law students are typically not very happy, but law schools appear to do a decent job of producing a lot of skilled lawyers. Of course, law schools are extremely selectionist, admitting students solely on GPA/LSAT (along with a modest amount of affirmative action).

Link is broken

The whole d**n system is broken!!

Leave it to Alex to criticize everything about the education system and then criticize it for attempting to innovate.

This hardly seems like much in the way of criticizing. He just seems to be pointing out the results of a study.

Why can't direct instruction be combined with student-led instruction in the following manner:
1. Teacher directly instructs the class,
2. The students who pick it up the fastest become de facto teaching assistants
3. They then proceed with student-led instruction while the teacher focuses on directly instructing the students who are the slowest.

That's how I do it... with the caveat that the students need a minute to think about the problem before you teach them how to solve it, otherwise many won't follow the explanation.

From my own experience attempting to run student-led activities, in the sense of "four students at a table try to discover the quadratic formula", I would agree that these are disastrous at the postsecondary level. Note however that the paper's description of direct instruction includes "students practice skills frequently". This would seem to rule out several commenters' ideas about excessively long lectures, which also seem pretty useless to me for college mathematics, at least until students have a certain level of mathematical maturity.

I'd be much more interested in a study that tested alternatives I actually thought might have a chance: for instance, what is the educational value of a well moderated debate every two weeks in a history class, assuming a teacher-moderator with median competence?

Not surprising. In general, the students are presumably the least knowledgeable and skilled people in the room (presumable there was SOME point to training the teacher in both domain and instructional methods).

Maybe for some subjects, with certain (very good?) students. With very small student:instructor ratio (2:1 or 4:1), with instructors who are good at Socratic methods. The idealized Oxford tutorial.

"presumable there was SOME point to training the teacher in both domain and instructional methods)"
Do they do it often?

Direct instruction might be the best approach to learning for some students at some ages. Is it the best general approach? I find it highly unlikely that direct instruction would be the best way to teach Lucy Allais's interpretation of transcendental idealism to Alex Tabarrok.

See Education Realist for more: https://educationrealist.wordpress.com/2018/03/28/four-obvious-objections-to-direct-instruction/

Don't confuse direct instruction (lower case letters), teaching directly to students, with Direct Instruction (capital letters), Zig Engelmann's scripted step-by-step program mostly for teaching basic skills. The Ed Realistic post is about Engelmann's DI. This post seems to be about generic direct instruction.

Ah, OK. Thanks!

Actually, I think you're wrong, Roger Sweeny. In Tabarrok's first sentence, he writes, "A slew of research shows that direct instruction produces superior results compared to other instructional methods." If you click on the words, "slew of research", it takes you to a link singing the praises of DI. So I think my original comment was relevant, after all.

At least for math teaching, students have remarkably consistent opinions over who is good and who is not. Probably for all the hard sciences too?

In my experience university students mainly agree on who are the unusually good teachers and always agree on who are the bad ones.

From reading the "Rate My Professors" website, I've noticed that you can tell a lot from the individual prose style of the comments.

5 stars: "Professor Jones really made me reassess my prejudice against the Austro-Hungarian Empire with his in-depth analysis of its emphasis on bureaucratic competence and top-down reform."

1 star: "Jonesy is boreing. LOL."

Perhaps my university experience was characterised by a narrower range of student ability than that represented on the website.

"Student-Led Classrooms Waste Teacher Skill" except that's not true if the adult in the room is not a teacher, is it?

In any case, if this works for some students but not others then the trick would get the students to self-select for the abilities required to do well in such an environment.

One might say the same of online instruction: those who are motivated and disciplined can make good use of it; goof-offs need not apply.

American education is structuring itself into an upside-down meritocracy, where those with academic ability, motivation, and discipline can obtain a high quality education at low cost, whereas obtaining even incrementally better outcomes for those who lack these qualities becomes increasingly costly.

Blow it all up. Sudbury schools populated with students who flunked out of regular schools have no teachers, just good outcomes.

Wow, there's all sorts of ambiguity here.

Like Roger Sweeney, I assumed that Tabarrok was talking about "direct instruction" (small d, small i), which is no one particular method, but an approach that is probably better called "teacher centered". That is, the teacher drives the activities in the classroom, whether it's a lecture or an activity, with the goal of explicitly teaching the students a particular task.

Then someone points out that the links go to Engelmann's Direction Instruction (capitals D and I). But I think that's an error by the author. If it is DI, then the whole article doesn't make sense.

Of course, the whole piece is screwy anyway, because almost no schools use "Student Centered Instruction". Very, very few schools let students run things.

If instead the author means that direct instruction works better than discovery, in which the students explore a concept and come to their own conclusions, that's both unsurprising and ancient news. I mean, why bother?

But then, the whole debate is kind of silly. The real issue, as you can see in the concepts, is rather how long do you lecture? How do you pace instruction? What do we do with the understanding that kids learning more abstract math (algebra plus) tend to need conceptual understanding, while younger than that do better with procedural?

That's where the action is. This piece is incoherent and uninformed.

"Why can't direct instruction be combined with student-led instruction in the following manner:
1. Teacher directly instructs the class,
2. The students who pick it up the fastest become de facto teaching assistants
3. They then proceed with student-led instruction while the teacher focuses on directly instructing the students who are the slowest."

Because parents don't send their kids to school to be teachers, and because just because kids *think* they know it doesn't mean they do and besides, the only people who think this is a good idea are people who get an undue amount of pleasure in thinking they're the ones who'd be selected to be the assistant teachers.

If you don't read any other comments, read this one.

If you don't read any other comments, read this one.

How is the student learning measured? As a teacher direct instruction would be helpful to guide students to the test that you designed. But that ultimately isn't how learning exists outside the classroom.

Comments for this post are closed