Direct Instruction: Yay or Nay?


Direct Instruction (DI) was developed by behaviourists Siegfried (Zig) Engelmann and Wesley Becker in 1967. DI does not address individual differences in children; it focuses on communicating concepts clearly enough for any and every child to understand. The DI method begins with letter sounds and then words; words are practiced many times before they are seen in context. The focus is on repetition, and for this reason it is seen as being quite authoritarian and inflexible. Cognitive researchers don’t like it because it was developed by a behavourist. Others favour its back-to-basics appeal; according to Principal Thaddeus Lott of Wesley Elementary in Houston, Texas, “It’s nothin’ but old-time teachin’ in a box.”

A component of the No Child Left Behind Act of 2001 was the Reading First program, which provides money for “Scientifically Based Reading Research” (SBRR) and for hiring “coaches” to impart the latest SBRR strategies to teachers. In 2006, a government audit revealed the program was plagued with conflicts of interest and wilful mismanagement. Reading First director Chris Doherty gave preferential treatment to states that used the DI approach. Review panels were sometimes stacked with DI backers or those with professional connections to the program.

Then there’s the passages from Zig Engelmann’s book, “The Outrage of Project Follow Through: 5 Million Failed Kids Later”:

The senior reading teacher and guru in one of our schools instigated an argument with me about reading – what it was, and how best to teach it. In the best cocktail-party style, we were polite, and the small group surrounding us was intent. The teacher’s premise was that the creativeness of teachers should not be trammelled by a lockstep program, like DI. She was well read, and quoted the literature with flourish. After the discussion went on for possibly ten minutes, one of our first-year teachers from the same school interrupted and ended the argument.

She said, “Angie, you know more about reading than I’ll ever know. You know linguistics, and all those theories I don’t understand. All I know how to do is follow the program. I do what it tells me to do in black type, and I say what it tells me to say in red type. But Angie, my kids read better than your kids, and you know it.”

I’m worried I have an anti-DI bias because I don’t like NCLB and I don’t like corrupt republicans. So I feel like I need to look into it further to figure out two things:

  1. whether DI is actually in conflict with cognitive research, and
  2. whether DI is as successful as its backers claim.

So I found a book called Taking Sides: Clashing Views on Controversial Issues in Early Childhood Education. Its description: “This debate-style reader is designed to introduce students to current controversies in early childhood education. The readings, which represent the argument of leading child behaviourists and social commentators, reflect a variety of viewpoints, are presented in pro/con format.”

And then I take a look at the publisher: McGraw-Hill. Wonderful. Don’t they publish a little thing called Direct Instruction? Yeah. Yeah, maybe I won’t take their word for it.

UPDATE, 2:36pm: A HA!!! It looks like it’s exactly in parallel with most cognitive research. I believe most of the backlash to DI is from bad teachers. From what I’ve read, DI is described as being “for students who can’t learn or teachers who can’t teach.” So having DI forced on you is like being told you’re an awful teacher. Well, if it walks like a duck and reads like a duck…

Comments 11

  1. Fraser wrote:

    I don’t know anything about any of this stuff, but shouldn’t state-institutionalized teaching methods take into account the average teacher’s abilities? Maybe the average teacher needs the dot’s connected for them. I guess the question is how good is the average teacher’s ability to teach something like reading? Can the average person, with an unrelated degree (so not a degree in “teaching to read”) teach someone else how to read? I don’t know the answer to that one, but in your post you seem to indicate that DI is an inferior way to teach because good teachers can teach better. But what about average and bad teachers? They certainly exist.

    There’s also a cost/benefit aspect to the problem. Maybe DI is a crappier way to teach, but I also bet it’s far more cost effective. And one might argue that people who want to learn to read should have to do some of the work on their own; work on their own that by default would be tailored to their specific needs.

    But let me emphasise that I don’t know anything about this stuff.

    Posted 23 Feb 2007 at 6:50 pm
  2. Fraser wrote:

    Ooh! There’s a rabbit next to my name.

    And if I’m expected to vote, I guess I’d vote yay, just for balance and because I already know how to read.

    Posted 23 Feb 2007 at 6:52 pm
  3. jason wrote:


    Be careful when you throw the word “average” around when talking about education. It’s not the “average” kids about whom I spend most of my time thinking, working and worrying; it’s the kids who aren’t average.

    If “any reader” could teach most kids to read, why is all this research being done? It certainly does get a lot of attention in education circles, and this spils over into politics (with most of the psycho-Right favouring initiatives like Direct Instruction and “back to basics” measures, despite quite a bit of research showing how ineffective both of those approaches are).

    Also, how do you define an “average” or “bad” teacher? This is a sticky issue. You can’t just look at the results from classes — marks, standardized tests, etc. — because of a host of reasons which would take me all day to write out.

    So, in conclusion, a complicated subject. But since educational approaches that mirror DI are universally awful (but, paradoxically, politically popular), I’ll have to come down against it. Susan Ohanian’s excellent One Size Fits Few would be a good place to start if you want to find out more about these sorts of issues.

    Posted 24 Feb 2007 at 11:24 am
  4. Fraser wrote:

    I never mentioned average kids, you can’t control what the average student is with the government, so I was only mentioning average teachers. But I will mention average students here.

    Any one style of teaching is going to be an advantage to some students and a disadvantage to others. That’s not to say that they are all equal, some styles will be inherently better than others I am sure.

    What if you have a mainstream style of learning but are being taught in a style designed to be accessible to the weakest student? Maybe you’ll learn anyway, but you’ll probably end up ripping your hair out because school is so boring. This problem would be worse if you were a strong student in that subject. This problem would also exist the other way around, weaker students being taught at a level of the strongest would also be bored with school.

    I guess I kinda came to the point, there needs to be a spectrum of education in order for everyone to learn. I get that, but what I don’t agree that the solution is something other than DI. Do we expect all of a person’s learning to come from the state? Isn’t it reasonable to expect someone to do that on their own? So the state makes you literate and numerate, but those who are really smart will go and develop their reading and math skills on their own. I remember my friends in highschool were splicing wires and hooking up the power-glove to their computers to control 3D imaging softwear. They didn’t learn to do that in school.

    I guess the two sides just looks at the role of the state’s education system. Is its purpose to make great readers? I’d suggest, because I’m intentionally arguing the other side, no. I’d suggest that the state’s goal is to give everyone a serviceable amount of education. It sounds horrible and draconian, but it really isn’t. There is a fixed amount of money we can spend on social services. And I think if people live in an environment that they are safe, free, happy and employed they will learn no matter what.

    One of the best indicators of a child’s education success is wealth. One might suggest that if parents earned more money their children’s test scores would go up.

    I’m not a psychologist so I don’t know if one causes the other. This is truly just “shooting from the hip”. And I’ll say again that I’m just opening this up for the sake of argument because I like to write my opinion on the internet. : )

    Posted 24 Feb 2007 at 1:06 pm
  5. Eve wrote:

    The most frustrating thing about the debate is that the effectiveness of Direct Instruction is greatly supported by research. It’s based on shitty theory, but because it has strong implementation guidelines, there’s not much for a teacher to fuck up. The ambiguity of implementation for good programs is what causes them to lose out. Fortunately, most educators (and I’m talking mostly principals and administrators) understand that and usually discount DI as a feasible program.

    Posted 25 Feb 2007 at 2:55 pm
  6. Eve wrote:

    Side note: DI is mostly prescribed for kids who are on the lower end of the scale. And seeing as I’ve heard tales of bad teachers inflating grades and sending kids off to the next grade without learning a whit in a single year, teaching to the lowest common denominator may sometimes be the only thing you can do.

    Having said that, DI is shit and I hope it’s replaced as the darling of researchers who don’t pay attention to the shaky methods of the papers they’re using to defend their arguments. Everyone knows the best way to teach is with the strap! Bring back the strap!

    Posted 25 Feb 2007 at 3:02 pm
  7. Fraser wrote:

    How can you attack a method in the same breath as saying that it has great tested results? Maybe I’m misunderstanding. Unless the tests are flawed and don’t actually measure learning.

    Individual police officers always have a solution to crime. But can you imagine a world where individual officers were each doing what they thought was “best” to do their job. They are on the front-lines too, they consider themselves professional and say things like “if only my ideas were implemented we wouldn’t be wasting all our time doing paperwork!” I’m glad that cops have a rigidly defined code of conduct. I’m glad they are forced to do paperwork that they consider a waste of time.

    Isn’t it natural for people to think that way? We are all smart and we all come to our own conclusions before we accurately test out an idea. But until we test something we don’t really know, do we? I think that if you got a large group of teachers together and they were all simply told to teach as best they could they would focus on teaching the things that they like to teach and neglect to teach things they consider chores. Some people don’t understand good grammar. I’m one of them, if I were to teach someone how to read I’d dive straight into sentence structure, readibility, crafting a story and stylistic techniques. There’s a double whammy here in my perception of my own teaching ability. 1)I’m teaching the things that are important to me. 2) When I see results in my students that reflect the elements I like and that I’m teaching I feel as though I am making a difference in their learning. And I am making a difference, I am teaching them things like readibility and style, but I am doing them a bigger dis-service by not teaching them grammar.

    There’s actually a third whammy. If I see a student’s writing that is weak on grammar I wouldn’t consider it all that important (because grammar isn’t that important to me). And because I’m not naturally that good at teaching grammar the likelyhood is that I would run into a lot of weak grammar. Once that happens it’s systemic. But I’d feel like a good teacher, I’d have many students that were great at the things that were I like.

    If I had a strict guide that forced me to teach all elements of writing equally I wouldn’t see the high levels of readibility and style, but I also wouldn’t ignore weak grammar. A writer who’s average in all areas is far better than one that has weak grammar. A writer with weak grammar will never get a writing job.

    That’s why we test things. That’s why we run experiments. A frontline person can’t see their own faults, so how could they see their faults in others? How could they possibly correct them? If DI has proven that it works better than other methods in a robust, peer-reviewed, controlled test and we can trust that the test actually measures learning then the only reasonable course of action is to use it.

    If DI has proven itself as effective in tests and you disagree that DI is the best thing for students you either have to attack the way the test was implemented (by saying that it doesn’t actually measure what we think it measures) or attack the practicallity of implementing the test in the real world (by saying it’s too expensive or something).

    But “researchers have spent too much time researching” is an unreasonable and illogical line of argument and often employed by all kinds of astrologists, mystics, psychics and all sorts of other bullshitters. Same with “I feel in my gut that my way is better,” prove that your way is better in a lab.

    Posted 26 Feb 2007 at 12:23 pm
  8. Fraser wrote:

    I really wish I had the ability to edit my comments.

    Please ignore the shift of reading to writing as the example I was using. Also please forgive the minor mistakes that happens when one writes on a computer (extra words that remain after a re-worked sentence, etc..).

    Posted 26 Feb 2007 at 12:38 pm
  9. Eve wrote:

    Yes, the tests are flawed. The main reason that DI succeeds is because it’s scripted, whereas many other strategies are not. Implementation is spotty for those methods than may have better standing theoretically but don’t have scripts that force teachers to teach to the method perfectly. Implementation is perfect for DI, so DI wins out.

    Posted 26 Feb 2007 at 3:15 pm
  10. Fraser wrote:

    You didn’t really say why the test was flawed. All you said that DI has perfect implementation, which seems to be it’s whole raison d’etre.

    Is a “spotty” system better than a one that is 100% consistent? Not normally. Would you go to war with a gun that ‘usually’ was accurate to 100 yards, but sometimes totally shot sideways or a gun that always was accurate to 75 yards?

    Consistency is worth a lot all on it’s own.

    Posted 26 Feb 2007 at 6:02 pm
  11. Eve wrote:

    That’s a reason to develop a better scripted system that uses good theory, not to just accept any ol’ scripted system. My argument is that the success of DI is being used to prove that DI (phonics based) strategies work better than whole word strategies, even though that’s patently untrue. DI’s success proves that DI’s delivery protocol is better, not that its instructional methods are better.

    Posted 26 Feb 2007 at 6:47 pm