I think this post helps to explain two phenomena about Direct Instruction: first, why it never expanded into secondary-level content, and second, why so many teachers feel instinctually averse to the DI approach. At higher levels (and in content-rich classes like history and science, where DI has never done much at all), the questions about what to teach and the connections between ideas are much more important, which makes DI programs much harder to write. DI has often been accused of being rote learning. The Project Follow Through evidence contradicts this, but we should remember Project Follow Through was only done in grades K-3. You can get great results on problem-solving measures in those grades just be being the best at teaching kids to read! Qualms about rote learning point to a real problem that you are highlighting. DI managed to sidestep much of that by focusing on the places where, as you say, decomposition is the most straightforward.
As the co-author of *Theory of Instruction*, I want to thank Becky for taking the time to read what is, truly, a very dense book. I will limit my comments to the quote about desirable outcomes for students: “They develop flexible understanding - mental models that allow them to navigate unfamiliar situations, see multiple routes through a problem, and adapt when the standard procedure doesn’t fit.”
and the comment about Zig: “He is …not so much a theorist of knowledge or curriculum design.
My response is to refer the interested reader to two publications that directly address these two comments:
This is an interesting post and has given me food for thought. I’m relatively new to Engelmann but have found his ideas helpful in planning my secondary math instruction. I think having CCSS standards helps remove some of the upstream issues discussed here which not all subjects benefit from. There is space in Engelmann’s framework to teach students to discriminate between problems that seem similar on the surface, but you’re right that more evidence is needed to see if there are benefits in less routine situations in higher grades. However, with so many students below grade level, I would be happy if Engelmann could solve the problem of rote fluency first, and we can figure out transfer to non routine problem solving later once we at least have the foundation covered! I would rather my students have an oversimplified definition of something like economic depression and be able to distinguish the most agreed upon features than too fuzzy of an understanding and poor discrimination that does not lead to the long term interest and flexibility of thinking that was intended. Focus on nuance and varied boundaries too early can easily become a constructivist approach where too much variance is considered valid.
I'm currently reading Kurt Angelmann's Direct Instruction, A Practitioner's Handbook. I'm finding it useful for a deeper dive into DI even though I'm not currently in the classroom. My school didn't use DI but I managed to get away with doing CMC math for my last few years in grades 1 and 2. Now I create curriculum for teaching writing.
One of the things I've learned more about from Kurt's book is the DI approach to history. He says "Students learn that historical actors have options to choose from in response to problems - accommodating, dominating, moving, inventing, or tolerating the problem. Students learn to apply this ADMIT model to past events, identify which option historical actors chose, and determine the consequences these choices had for future events...". A good teacher would lead a lively discussion leading to further questions. A less skilled teacher would insist that there was a correct answer and it had to be one of the above. DI is opening the door with, in my view, an exciting way in to a discussion. I think the role of DI here would be to give enough knowledge and ask the right questions that would enable students to go to that edge where they can debate and explore gray areas.
In this case I would say that DI has done the hard work before instruction. It's the teacher's job to do the work after instruction.
I love all the comments. What I most wanted to comment about is that most educators that I have ever met, and I’ve been around, have no idea who Engelmann is or what DI is. Further direct instruction is not thought of highly.
Thanks for this thoughtful discussion. Something I've been confused about when I've watched Engelmann videos is, when was the initial learning done before children were comfortable enough to apply it in these call and response contacts? Or is the idea that kids ambiently absorb context as their peers shout out answers?
In my teaching, I've often had the false impression that nearly all students in the room understand something, because I'm getting a variety of students shouting out correct answers. Even in Engelmann's classic arithmetic film that's on YouTube, I notice that there are some kids who seldom call out the answers. Isn't there a need for a shift in modality sometimes, so that kids have the chance to explore these concepts with some amount of constructivist agency?
Very interesting. Thank you. I wrote a blog arguing for a normative interpretation of Engelmann, which, entails that what Engelmann is doing is ‘training’ students to be able to do certain things - not trying to induce ideas into their heads.
I had a little discussion with David Didau about why that matters (the link to the blog is in this post too):
I think it’s particularly problematic for philosophy because the whole crux of DI is that you try to remove the fuzziness in concepts - whereas in some subjects, especially philosophy, epistemological diffusion IS the point! The fact we disagree about the characteristics of a concept is a feature not a bug.
Yes, I think that's absolutely right. Although, I suppose, we still need to have agreement about certain concepts in order to make disagreement possible. E.g. We might disagree about whether an argument is circular or not, but we have to have agreement about how we use the concept of 'circularity'. Wittgenstein had a nice notion - he called them 'hinge concepts' - concepts which stay fixed so the door can swing. I think that's where Engelmann's approaches are v useful - establishing agreement on things like that, but after that, we need something else. As you say, the point of philosophy, law, politics, history, is that we are trying to establish agreement as to how apply our concepts to new and fringe situations. They're not laid out in advance.
Yes, and Becky gets further into that in a really helpful way too. Essentially there’s a lot of decomposing to do in subjects which don’t start with basic building blocks.
I wish I could have read this when I first joined Substack, as it would have been very useful to help navigate the various opinions about direct instruction that I find here. I think this hits the nail on the head and is part of what I found frustrating about reading the evangelists on substack .
I’ve found myself in many arguments with DI proselytizers where I’ve pointed out that it is almost impossible to teach strong mathematical problem-solving habits (at the upper middle to high school level) through direct instruction alone. Direct instruction is very helpful for getting students to master procedures and identify relationships but it does not help them approach novel problem-solving situations. I look forward to reading your follow-up post on this topic!
Love this. That opening paragraph is 👌. Tom Needham is about to publish Engelmann In Action - a great effort at making sense of it - and this issue of who designs the curriculum sequence was something we discussed. It seems a lot harder outside Maths.
You seem to be writing that people cannot learn some thing by direct communication because the things are fuzzy and so we can’t communicate what the thing is.
It seems you are attempting exactly what you describe is impossible. As with games or fish it’s not hard to explain some nouns are difficult to define.
Why would it be hard to explain that experts don’t agree on the exact definition of a depression?
I think this post helps to explain two phenomena about Direct Instruction: first, why it never expanded into secondary-level content, and second, why so many teachers feel instinctually averse to the DI approach. At higher levels (and in content-rich classes like history and science, where DI has never done much at all), the questions about what to teach and the connections between ideas are much more important, which makes DI programs much harder to write. DI has often been accused of being rote learning. The Project Follow Through evidence contradicts this, but we should remember Project Follow Through was only done in grades K-3. You can get great results on problem-solving measures in those grades just be being the best at teaching kids to read! Qualms about rote learning point to a real problem that you are highlighting. DI managed to sidestep much of that by focusing on the places where, as you say, decomposition is the most straightforward.
Great food for thought, thank you for writing!
As the co-author of *Theory of Instruction*, I want to thank Becky for taking the time to read what is, truly, a very dense book. I will limit my comments to the quote about desirable outcomes for students: “They develop flexible understanding - mental models that allow them to navigate unfamiliar situations, see multiple routes through a problem, and adapt when the standard procedure doesn’t fit.”
and the comment about Zig: “He is …not so much a theorist of knowledge or curriculum design.
My response is to refer the interested reader to two publications that directly address these two comments:
My primary recommendation for the interested reader is "Designing Effective U.S. History Curricula for All Students." The reader will need to click "View full text" https://www.researchgate.net/publication/319664975_Designing_Effective_US_History_Curricula_for_All_Students#fullTextFileContent
It provides about a dozen "mental models that allow them to navigate unfamiliar situations."
Unfortunately, Substack does not allow me to attach some of those mental models.
Second, a few chapters of interest can be found in "Higher Order Thinking: Designing Curriculum for Mainstreamed Students." Some ot the chapter titles are The Search for a Unified Social Studies Curriculum: Does History Really Repeat Itself, Problem Solving and Transfer, and The Fundamental Skills of Higher Order Thinking https://www.researchgate.net/profile/Doug-Carnine/publication/371950105_HigherOrderThinking/links/649db4208de7ed28ba649504/HigherOrderThinking.pdf
Thank you for sharing!
This is an interesting post and has given me food for thought. I’m relatively new to Engelmann but have found his ideas helpful in planning my secondary math instruction. I think having CCSS standards helps remove some of the upstream issues discussed here which not all subjects benefit from. There is space in Engelmann’s framework to teach students to discriminate between problems that seem similar on the surface, but you’re right that more evidence is needed to see if there are benefits in less routine situations in higher grades. However, with so many students below grade level, I would be happy if Engelmann could solve the problem of rote fluency first, and we can figure out transfer to non routine problem solving later once we at least have the foundation covered! I would rather my students have an oversimplified definition of something like economic depression and be able to distinguish the most agreed upon features than too fuzzy of an understanding and poor discrimination that does not lead to the long term interest and flexibility of thinking that was intended. Focus on nuance and varied boundaries too early can easily become a constructivist approach where too much variance is considered valid.
I'm currently reading Kurt Angelmann's Direct Instruction, A Practitioner's Handbook. I'm finding it useful for a deeper dive into DI even though I'm not currently in the classroom. My school didn't use DI but I managed to get away with doing CMC math for my last few years in grades 1 and 2. Now I create curriculum for teaching writing.
One of the things I've learned more about from Kurt's book is the DI approach to history. He says "Students learn that historical actors have options to choose from in response to problems - accommodating, dominating, moving, inventing, or tolerating the problem. Students learn to apply this ADMIT model to past events, identify which option historical actors chose, and determine the consequences these choices had for future events...". A good teacher would lead a lively discussion leading to further questions. A less skilled teacher would insist that there was a correct answer and it had to be one of the above. DI is opening the door with, in my view, an exciting way in to a discussion. I think the role of DI here would be to give enough knowledge and ask the right questions that would enable students to go to that edge where they can debate and explore gray areas.
In this case I would say that DI has done the hard work before instruction. It's the teacher's job to do the work after instruction.
I love all the comments. What I most wanted to comment about is that most educators that I have ever met, and I’ve been around, have no idea who Engelmann is or what DI is. Further direct instruction is not thought of highly.
Thanks for this thoughtful discussion. Something I've been confused about when I've watched Engelmann videos is, when was the initial learning done before children were comfortable enough to apply it in these call and response contacts? Or is the idea that kids ambiently absorb context as their peers shout out answers?
In my teaching, I've often had the false impression that nearly all students in the room understand something, because I'm getting a variety of students shouting out correct answers. Even in Engelmann's classic arithmetic film that's on YouTube, I notice that there are some kids who seldom call out the answers. Isn't there a need for a shift in modality sometimes, so that kids have the chance to explore these concepts with some amount of constructivist agency?
This is an exceptional post and gets to the heart of the challenges with general instructional theories.
I really enjoyed your use of downstream too. It helps me situated where instruction relates to the rest of the educational field.
It feels like it’s time to return to the knowledge question in education.
I find the work of Jan Derry and Michael Young at the IoE particularly revealing when exploring the knowledge question.
Very interesting. Thank you. I wrote a blog arguing for a normative interpretation of Engelmann, which, entails that what Engelmann is doing is ‘training’ students to be able to do certain things - not trying to induce ideas into their heads.
I had a little discussion with David Didau about why that matters (the link to the blog is in this post too):
https://substack.com/@daviddidau/note/c-202066681?r=35t6uv
I think it’s particularly problematic for philosophy because the whole crux of DI is that you try to remove the fuzziness in concepts - whereas in some subjects, especially philosophy, epistemological diffusion IS the point! The fact we disagree about the characteristics of a concept is a feature not a bug.
Yes, I think that's absolutely right. Although, I suppose, we still need to have agreement about certain concepts in order to make disagreement possible. E.g. We might disagree about whether an argument is circular or not, but we have to have agreement about how we use the concept of 'circularity'. Wittgenstein had a nice notion - he called them 'hinge concepts' - concepts which stay fixed so the door can swing. I think that's where Engelmann's approaches are v useful - establishing agreement on things like that, but after that, we need something else. As you say, the point of philosophy, law, politics, history, is that we are trying to establish agreement as to how apply our concepts to new and fringe situations. They're not laid out in advance.
Yes, and Becky gets further into that in a really helpful way too. Essentially there’s a lot of decomposing to do in subjects which don’t start with basic building blocks.
I wish I could have read this when I first joined Substack, as it would have been very useful to help navigate the various opinions about direct instruction that I find here. I think this hits the nail on the head and is part of what I found frustrating about reading the evangelists on substack .
I’ve found myself in many arguments with DI proselytizers where I’ve pointed out that it is almost impossible to teach strong mathematical problem-solving habits (at the upper middle to high school level) through direct instruction alone. Direct instruction is very helpful for getting students to master procedures and identify relationships but it does not help them approach novel problem-solving situations. I look forward to reading your follow-up post on this topic!
Love this. That opening paragraph is 👌. Tom Needham is about to publish Engelmann In Action - a great effort at making sense of it - and this issue of who designs the curriculum sequence was something we discussed. It seems a lot harder outside Maths.
You seem to be writing that people cannot learn some thing by direct communication because the things are fuzzy and so we can’t communicate what the thing is.
It seems you are attempting exactly what you describe is impossible. As with games or fish it’s not hard to explain some nouns are difficult to define.
Why would it be hard to explain that experts don’t agree on the exact definition of a depression?