Educational approach

Marginal Revolution has an interesting discussion on teaching approach centered on Direct Instruction, which is generally thought of as script-focused educational approach.

Direct Instruction is more than just following a script. The scripts for DI are engineered using a build/test/fix cycle to have the least ambiguity possible. For instance experiments were performed on how to teach a young child the color red. Do you show them a red object first and say red? or do you show them a non-red object first and say "this is not red?" They created several 10 step scripts using the same 10 steps just in different orders, and the best one would teach a child the concept in 20 instances, the worst in 60 instances. Now magnify this across the thousands of concepts a child must learn to be educated.

The scripts are designed to create many small interactions where the students are tested for comprehension by getting their response. A small chunk of information is presented, and then the students are immediately tested to see if they understood, and correction is given right away. The students stay on topics until mastery. This sometimes requires slowing them down, and other times speeding them up. To do this the students must be placed with other students at the same learning place and learning pace. This is not regarded as a static placement, and students will have to be re-evaluated and moved to different groups on a regular basis.

DI not only tracks the individual student it makes schedules for teachers and tracks them to those schedules. If the teacher falls behind, then they try and figure out what is going wrong. Video's of the teachers teaching are made for review to try and anticipate problems.

The linked discussion is more interesting that anything I'll say here, but here's (more than) a couple of call outs:

• When Meryl Streep shows up to make a movie, they hand her a script. But when a new teacher shows up to teach her first class, in many school districts they ask her to invent her own lesson plan. What gives?
• Isn't DI what the military has used during and since WWII for their instruction?
• This is basically the argument that Atul Gawande makes about Cesarean sections in the relevant chapter of Better ("The Score.") Obstetricians used to have all sorts of clever and fancy forceps techniques for delivering babies in various difficult positions. But those techniques required a lot of reading to learn; even more practice to master; and a natural level of spatial intuition, coordination, and judgment that not every doctor had. Shifting to routine C-sections may have sacrificed the ability of truly outstanding obstetricians to deliver even misaligned babies without surgery, but it gave average obstetricians a single standard script they could follow to make most deliveries a matter of routine.
• The complaint that i hear from teachers is that their teaching is TOO scripted. They say that they have no freedom to teach the way that they would like. Perhaps the problem is that the same script doesn't work with every child. For example some children don't do well, just because of the (imho) crazy demand that little kids sit quietly in desks for hours at a time.
• That's the same crowd who hate NCLB. Hate to actually show their students know how to read, write or add. "Too scripted" or "having to tech the test" sound a bit like accountability to me. We'd hate to have that!
• Where does this leave the uber teachers (and their fortunate students)? If the method brings up the average success, it's good from a utilitarian standpoint; but would there be a way of maintaining those classrooms which are exceptional? I don't think those teachers would care much for the script.
• The problem with a script is that nobody can agree on it. There is already a standard curriculum for public education, which is driven by standardized testing. Not to mention, it's boring and doesn't encourage critical thinking. The learning experience isn't one-way, teacher-to-student. It's a mutual effort that requires desire from all stakeholders (teachers, parents, and the students themselves). The problem among the low-scoring population is that the stakeholders don't feel invested.
• I'm a science teacher at the most diverse high school in our county. Our standardized test scores have been consistently below those of the other county schools, with the difference written off as an inevitable result of teaching a "much more challenging population". Last year, one of our science disciplines went to a more scripted approach, and we soundly beat the other schools."Scripted" in this case means having a clear, common set of objectives, required knowledge, assessments, materials, and activities. The actual instruction method is up to the teacher. When the goals are clear, there is far less opportunity for teachers to drop the ball on covering vital topics. There's much more to it, but the "scripted" nature is a crucial aspect. An additional benefit is that new teachers no longer have to develop their lessons from whole cloth, formerly a cause of "one year and out" teachers.
• Teaching is a kind of performance art and it shouldn't be surprising that working from a good script can really help. Of course it all depends on the quality of the script. There is already a market in textbooks and lesson plans, and a large number of different approaches to teaching. Maybe the question should be why this market isn't more efficient?
  
• The results of a study can be no more accurate than the standardized test used to measure the results, and such test are more accurate when testing rote learning. I have seen DI used to teach reading in early grades and it works well with most students but not all. But reading is a rote learning as is multiplication tables, but many subjects are taught for the reasoning skills they impart not the facts. Since the ability to read is needed to do well on most standardized test one would expect a correlation between a technique that teaches reading effectively any other skill tested.
• Our educational system suffers from one major problem and it's not using the wrong pedagogical techniques. It's that the middle class doesn't want to pay for the education of the lower classes. They enforce this desire by using local property taxes to fund education, thus ensuring that the poorest school districts are also the ones with the least amount of money.
• The description of direct instruction reminded me of the book we used to teach our children to read, "Teach Your Child to Read in 100 Easy Lessons." It worked incredibly well for two of our children, who learned to read at late age 3, and age 4, respectively.
  
• Teaching from a script is not the same as an actor reading from a script. This is because students make errors which need to be corrected, preferably immediately. When a student makes a mistake, the teacher must stop the presentation, evaluate the mistake, determine the required remedy, implement the correction, test the student to see if the remedy was successful, and then resume the presentation. This is one reason why DI is more effective than most other method of teaching. DI approaches education as an engineering problem. The DI "script" is designed to minimize student errors, acquire a high rate of feedback from students to determine if errors are present in the student learning, and to remedy those errors as soon as they are detected. The motto is: the teacher hasn't taught, if the student hasn't learned.

My initial though, based entirely on the the MR post is that it seems to be a great program for process topics (reading, math, some sciences) but perhaps less useful for critical thinking topics, though I could easily be wrong. I do find it disconcerting that new teachers must formulate new lesson plans. I know this isn't the norm everywhere, but how can every school/district not have a base lesson plan in place for all subjects?