The libertarian side of the blogosphere is all abuzz about "Direct Instruction" at the moment, thanks to a Marignal Revolution post by Alex Tabarrok touting the method:
Ayres argues that large experimental studies have shown that the teaching method which works best is Direct Instruction (here and here are two non-academic discussions which summarizes much of the same academic evidence discussed in Ayres). In Direct Instruction the teacher follows a script, a carefully designed and evaluated script. As Ayres notes this is key:
DI is scalable. Its success isn't contingent on the personality of some uber-teacher....You don't need to be a genius to be an effective DI teacher. DI can be implemented in dozens upon dozens of classrooms with just ordinary teachers. You just need to be able to follow the script.
Contrary to what you might think, the data also show that DI does not impede creativity or self-esteem. The education establishment, however, hates DI because it is a threat to the power and prestige of teaching, they prefer the model of teacher as hero.
Jim Henley also takes this up, and raises a sensible market-based objection to the idea:
Direct Instruction fails for different sociological reasons. Star teachers don't solve the problem of education because not enough teachers are capable of stardom. Direct Instruction doesn't solve the problem because not enough teachers are willing to do that work. Direct Instruction sounds like utter drudgery from the teacher's perspective. I'm inclined to think the resistance to it in the existing education establishment indicates not that the class of teachers are full of themselves, but that you could no more fill America's schools with people willing to speak the same script day in and year out than you could fill them with Jaime Escalante clones.
This is still essentially about the scripting issue, though, which seems to me to be missing the point. After all, if you look at the summary of Direct Instruction provided at the first of Tabarrok's links, the scripts are really only one part of the method, and frankly, they don't look all that important, relative to the other bits:
- Homogeneous Skill Grouping: Children are grouped according to their levels of ability, rather than according to age or other factors. If you are going to teach the same material to a group of children, they clearly benefit most if they are all able to follow the material.
- Scripted Class Sessions: Teachers use pre-designed scripts when teaching. The scripts are based on extensive research regarding student retention, and every aspect of every script is based upon results that were demonstrated through research. The great advantage of this approach is that every teacher using the script becomes the beneficiary of that research and will probably teach much more effectively than if left to his or her own devices.
- Intense, Constant Student Interaction: The scripted sessions consist primarily of sequences of stimulus/response pairings, wherein the teacher stimulates the class with a description of a concept, an illustration of the concept through an example, and finally a request that the class repeat the example. The class responds orally, usually as a group.
- Teaching to Mastery: The group does not move on until everyone in the group understands the material.
Frankly, the scripts look to me like the least important part of that. I suspect that if you had the "Homogenous Skill Grouping" and the "Intense, Constant Student Interaction" parts, you'd see big improvements regardless of whether everybody taught from the same script or not. I suspect that the only real effect of the scripts is to make it easier for some teachers to do "Intense, Constant Student Interaction," when their natural inclination would be more in the direction of non-interactive lectures.
As someone who teaches a lot of intro classes, I can say that there are few things more maddening than teaching wildly heterogenous classes. It really sucks to look out into the room and see that the weakest students are being completely blown away, while the best students are being bored out of their skulls. There really isn't a good way around that, either, in classes where you've got students with a huge range of abilities, unless you're naturally one of the "heroes" that Tabarrok thinks are overemphasized, and can put in extra hours either tutoring the weak students, or providing extra enrichment for the strong students.
If the whole class is weak by some objective standard, there are things you can do to deal with that-- step down the level of the lectures, increase the amount of time spent on practice problems and repetetive drill. If the whole class is strong by some objective standard, there are things you can do to adjust-- push the pace of the class, explore additional topics of interest. But when you've got a mix of really good and really weak students, all you can do in good conscience is to pitch the class somewhere in the middle, and hope that neither extreme is suffering too much.
Just the skill grouping by itself would make a substantial difference, I think. And "Intense" interaction among students and between students and faculty is well known to be an effective teaching tool. There are no end of studies showing that "Active Learning" techniques do a better job of teaching students physics concepts than traditional lectures do, and I have no doubt that extends across disciplines. That's why good humanities classes use a discussion format, after all, and it's a big part of why students are willing to spend $40K to attend small colleges.
As for the general issue of teaching from a script, I think it would really have to depend on the script. If it's really a minute-by-minute sort of script, then yes, it sounds absolutely stultifying. If it's more of a "you will teach these topics on these days, using these examples" sort of thing, then that might not be too bad. That's really not that different than what you face in teaching one section of a multi-section course.
It would depend a lot on the subject-- the example lessons on this page look pretty hideous to me, but that's largely because they're primary school material, and I'm not cut out to be a primary school teacher.
As with most education-related issues, there are a number of other elements about the arguments Tabarrok and other make that I find kind of dubious-- if someone suggested teaching, say, Introduction to Econometrics off a strict script, would he happily sacrifice his own "power and prestige" as a teacher? It's really easy to suggest these sorts of measures for other people, after all.
Really, though, I think the script thing is probably a red herring. At least based on the descriptions linked from the original post, I think other parts of the method are a lot more important to its success than the script business.
Education & Careers
Politics
Comments
# 1 | marciepooh | September 28, 2007 10:39 AM
I heard a story on NPR a couple of years ago about an elementary school that used what sounds, to me, like DI or at least a similar idea for teaching reading. They also applied some of the principles to other subjects and the whole day. One of the problems cited by teachers who had left the school was that when students didn't "get it" the teachers weren't allowed to bring in other resources. All they could do was hand the kids the same workbooks over and over. If the scripting is too specific and outside resources discouraged, then the kids for whom it doesn't work get left out.
They are, IMHO, pluses and minuses to skill grouping. In some areas, math for instance, grouping by skill level is wonderful; everybody learns better. But if kids are always separated into the "smart" and "dumb" groups then I think you lose some thing, too. I imagine humanities are probably easier to teach to a wide range of students than science and math.
# 2 | marciepooh | September 28, 2007 10:42 AM
There are, IMHO,... (sorry, I should really proof-read.)
# 3 | Chris Goedde | September 28, 2007 11:59 AM
On the homogeneous/heterogenous student population question, the only real solution is smaller student/teacher ratio so that heterogenous student populations can be broken down into smaller homogeneous groups. Somehow, I don't think the people promoting this will sign on to that remedy.
As for the other parts, they sound remarkably like the ideas coming out of the physics education community, e.g. McDermott's tutorial-based physics instruction. Also requires fairly low student/teacher ratio, of course.
# 4 | Jonathan Vos Post | September 28, 2007 12:34 PM
In a sense, any adequate teacher can teach the students who "get it." One of the great challenges is to find a way to get through to the students who "don't get it."
No panacea exists; no one-size-fits-all method. I find the scripting or "Direct Instruction" to be an appalling attack on the dignity of students and teachers alike.
There's a grain of truth in the notion: "Star teachers don't solve the problem."
My mother, my wife, and I have been "star teachers, and there aren't enough of us. True. The parallel is with the classic flaw in Management that heroic effort by an individual manager can solve systemic problems.
But the paradox of American education is that the best is as good as anywhere in the world, while the mean has fallen to third-world standards. So how best to acknowledge the heterogeneity of teachers, students, facilities, neighborhoods, principals, staff, teacher training institutions, parents, and all other stakeholders in the educational system?
And how do we get there from here?
Some countries have systems that do a better job than others. The heterogeneity of countries gives us data to analyze. "No Child Left Behind" is a Federal blow at heterogeneity of States within the USA. In my opinion, seeing it at work, it forces all school districts to dumb-down their teaching, increases "teaching to the test", and resulted in individual states suing the Federal government on the basis that they would have to lower their standards to comply with Federal guidelines. The Federal regulations, by the way, have no exit strategy. What is to be done when essentially no school districts reach the 100% comptenency in Math and English mandated by the Feds? I am nearly 100% sure that we'll see that. In Pasadena Unified School District, that goal can be achieved only with 9% improvement in test scores every year, starting now. The past 5 years in PUSD have seen modest trend upwards in test scores (I think because teachers teach to those tests) for White students, and stagnation for African-American and Hispanic students.
As George W. Bush said yesterday, urging renewal of what I consider the failed "No Child Left Behind" unfunded mandate: "Childrens do learn." That video clip was played on both David Letterman and Jay Leno.
Teaching students who "don't get it" -- 6 semesters in college and university, summer school in Pasadena -- made me sometimes weep, at home, after taking off the coat and tie. But EVERY one of those hundreds of students who tried (came to class, did homework, engaged me in one-on-one discussion) did finally "get it." Even if they'd flunked the same course before, or twice before, or even three times before.
No standardized script could have gotten through to these students. What heartless bastard would just write them off, throw them into the gutter, and enshrine it as part of Schumpeter's Creative Destruction of Capitalism?
# 5 | The New York City High School Math Teacher | September 28, 2007 5:09 PM
As a 6 year veteran of the inner city math classroom, what I would distrust is the competence of the authors of the scripts. The 2003 Prentice-Hall Math A/B curricula/scope&sequence/day to day schedule used in NYC public schools through this year comes to mind. Some real screamers in there - where were the proof-readers?! Of course, I have never used it.
Also, it would be killingly boring. I enjoy teaching math because I enjoy math, and I enjoy the opportunity to improvise and adapt teaching to the moment. If I had to submit to a litany prescribed by some inner-circle timeservers with patronage, damn and hell, I'd be out of there like a shot.
Homogenous grouping is totally outré politically, no matter that it works better. Of course, I don't understand why it is outré - anti-elitism? I have a section of 30 with seven distinct skill levels, all the way down to innumerate - is it an efficient use of my time to personalize instruction that precisely? Is there enough time at all?
Further, how is it possible to plan for the potential branchings of individual student interest and endeavor in such a scripting? I'll tell you - you can't.
The script won't contain the potential branches, and the instructors will be too incompetent to direct the students themselves, because the script to delivers them from having to think about the material in any structural way.
I underline this: scripting is done when you don't have confidence that your teachers are competent.
# 6 | guthrie | September 28, 2007 6:02 PM
To put it another more general way, a script cannot cover the variety of individuals and how they will each learn in specific circumstances. Whereas a properly experienced and well trained teacher, will be able to adapt the methods and styles for each individual child as well as the class as a whole. This is not possible with a script, unless it is almost infinitely variable, which I am sure would be an incredible piece of work. So incredible, that I am sure it has not and will not be done.
# 7 | guthrie | September 28, 2007 6:06 PM
high school maths teacher:
"I underline this: scripting is done when you don't have confidence that your teachers are competent."
Damn right. But then why do you want competent professionals? They'll only want more money, when instead you can install scripts into cheaper incompetents who can be replaced at short notice due to their interchangeability.
Now, the intelligent amongst us see a large difference between interchangeable parts in cars and machine tools, compared to children in schools. But it seems the great big machine grinds on and on, grinding people and their differences into dust.
(I just can't remember the specific long words to describe it all. TIme for me to go to bed)
# 8 | Coin | September 29, 2007 4:52 PM
Damn right. But then why do you want competent professionals? They'll only want more money, when instead you can install scripts into cheaper incompetents who can be replaced at short notice due to their interchangeability.
At this point, the whole script thing is starting to sound like the scripts used by outsourced customer-service call centers. Which raises the point, if they're not interacting with the child except as according to a script, why bother with having a physical person in the room? You could have the teaching done by someone teleconferencing from another continent. Or a computer simulation of a person. Or you could just let the kid read the damn script himself.
Chad suggests the scripts aren't the point and are kind of a red herring, but on my cursory understanding of all this it seems to me that it's hard to gloss over the scripts, considering that they are completely antithetical to every other aspect of the program. Most of the points of this DI thing seem to have the chief effect and benefit of targeting each student personally, rather than allowing them to get lost in the shuffle of a large classroom. Yet the script by its very nature (even if this does scale with how fine-grained the script is) is the antithesis of personalized instruction...
# 9 | PhysioProf | September 29, 2007 8:30 PM
"The class responds orally, usually as a group."
This scares the shit out of me. Are we teaching kids to think, or indoctrinating them into conformity?
# 10 | HennepinCountyLawyer | September 30, 2007 3:48 AM
Coin said: ". . .if they're not interacting with the child except as according to a script, why bother with having a physical person in the room?"
Many years ago I heard an address by Al Shanker (then president of the American Federation of Teachers) that touched on this point. He pointed out that if there's no interaction, for any given subject you can almost certainly find or produce a video that will do a better job than a teacher talking at the class. He suggested we should actually be making more use of videos (this was before there were computers in the classroom) for one-way instruction, and let the teachers work in small groups or one-on-one. The essence of teaching, he said, is not lecturing but coaching.
Of course, that gets us back to the uber-teacher. But the point remains: if we're just using scripts there's no need to have a live person reading it.
# 11 | Jon H | September 30, 2007 4:58 PM
So what happens if a student asks a question that is interesting, on topic, and deserves attention, but goes off the script?
It seems like the script would require that student to be squelched, in favor of continuing with the script - which can't help the interest levels of the students; why bother thinking?
It doesn't seem like genuine interaction to me if the students are just giving group choral responses as if reciting the Pledge.
# 12 | Jon H | September 30, 2007 4:59 PM
Come to think of it, this sounds like someone trying to turn Computer Based Training into a method of instruction by humans.
# 13 | Cabalamat | September 30, 2007 7:12 PM
A bit off at a tangent, but doesn't "No Child Left Behind" imply "No Child Allowed To Progress Faster Than The Slowest".
# 14 | Jonathan Vos Post | October 1, 2007 4:35 AM
Re #13: "... doesn't 'No Child Left Behind" imply 'No Child Allowed To Progress Faster Than The Slowest.'"
Not quite. They're institutionalizing what benefitted me in public schools of the 1950s and 1960s, after Sputnik. "Every Child who does well enough on the standardized tests gets to sit in the back of the room reading whatever he or she wants, and becoming an autodidact."
If we can't teach teachers how to teach subjects whose contents they don't know, we can at least teach teachers to leave children along to teach themselves. Only now, instead of letting them read the encyclopedia and old national geographics as I did, the children can mess around on computers.
Lacking leadership, the educational system is accidently achieving the libertarian objective of getting government out of the way, and letting people achieve what they are motivated to achieve, by their own devices, and by means of freedom of association with whomever can actually teach them, in the traditional Master-Apprentice learn-on-the-job system.
And thus, as George W. Bush stated: "Childrens do learn."
# 15 | The New York City High School Math Teacher | October 1, 2007 12:35 PM
I don't know so much about that, Jonathan.
I view it more as a malignant effort to Taylorize and de-skill the classroom. The effect is largely the same, but the moral undertone is entirely different.
By the way, there is something to be said for extremely extensive libraries of teaching materials, down to scripts for particularly successful lessons, annotated with informative notes, observations, and debriefs from teacher, student, and witnesses. You see, improvement of instruction doesn't come from standardizing output according to publisher pre-measured and pre-organized lessons, but from conscientious fine-tuning, derived from self-examination and skilled peer observation and comment. But there is no space for this*, no matter that it works and makes happier, more skilled, and more effective teachers.
*Aside from fake, one hour per month Kabuki shows.
# 16 | Jonathan Vos Post | October 1, 2007 6:42 PM
I agree with "The New York City High School Math Teacher" component of "malignant effort to Taylorize and de-skill the classroom."
This would allow administrators to treat teachers as fungible (i.e. completely interchangeable) and thus be able to pay them less, replace them more easily, spend less on training them, and decreasing the power of teachers' unions.
What I find reprehensible about this is the whole dehumanizing vision of teachers and students alike as assembly-line stations and products. Besides, of course, that it usually doesn't work.
Let's see. I'm spending about $1,500 to $2,000 per quarter, for tuition, fees, textbooks, gas, and parking, at Cal State L.A.'s Charter College of Education (a cheap school of ed) to receive Full Credentials to teach Math and Science in California Public Schools. That's roughly $9,000 to $12,000 out of pocket, on top of what I spent for two B.S. degrees from Caltech, a M.S. from UMass in Computers, an All-But-Degree PhD in computers, and an 95%-of-all-but-degree MFA in poetry.
So far, all I can be is a Substitute teacher for roughly $110/day. Break-even is thus roughly 82 to 109 days of substitute teaching, with the grad courses taken at night after the more-than-8-hour days. not including present value of money (i.e. interest rates).
Hence it takes a vow of poverty to be a public school teacher, compared to what one could earn (as I did) with the graduate degrees themselves before taking a deep paycut to help children and teenagers. So Taylorization adds insult to injury.