In 2016, I was sitting in the classroom of a Melbourne school as Dr. Kerry Hempenstall described the early stages of a reading program. He projected a series of letters on the screen. First, he displayed an “f.” “This is an ‘f’,” he said. Then he displayed an “f” written in a different font. “This is an ‘f’,” he said. Then he displayed a red “f.” “This is an ‘f’,” he repeated. Finally, he displayed a “p.” “This is not an ‘f’,” he said. Hempenstall explained that this last letter represented a “non-example.”
I grew excited. For a few years, I had been aware of research demonstrating that many of the misconceptions students have about mathematics stem from overgeneralizing a concept. In other words, they take a fact or rule that is true in one situation and apply it to a situation where it is not true. “Non-examples” could be the key to preventing such misconceptions from arising.
By now, you may be starting to think I am a sad individual. Who cares about these details? Isn’t education supposed to be about big ideas such as critical thinking or creativity? Nobody wants to go to an education conference at a flash convention center, eat tiny smoked salmon sandwiches, and hear someone talk about a “p” not being an “f.” They want inspiration, passion, and a bunch of platitudes about relationships.
The reading program Hempenstall was describing is an example of Direct Instruction. Confusingly, the term “direct instruction” has several different meanings and is often used as a shorthand for any kind of teaching where the teacher fully explains concepts and models how to complete tasks. However, the kind of Direct Instruction that Hempenstall was describing consists of a specific set of programs that have their origins in 1960s America and the work of the late Siegfried “Zig” Engelmann and his collaborators, first at the University of Illinois and then, from 1970, at the University of Oregon. To distinguish these programs from direct instruction more generally, a convention has developed where the “D” and “I” are capitalized.
Engelmann had little time for the grandiose ideas that were popular on the education conference circuit. “There is no big thing. It’s all pick, pick, picky details. Direct Instruction is just attention to a lot of tiny details,” he claimed.
If teachers know anything about Direct Instruction—an unpopular topic in teacher training courses—it is that the lessons are scripted. However, this was not the original plan. As with many aspects of these programs, scripts developed as a refinement in response to problems Engelmann faced during development. “Whenever [teachers] went to correct a mistake, they would get wordy or redundant. They would start appealing to things that the kid didn’t know,” he explained. It reminds me of the iterative process used by engineers.
Contrast Engelmann’s thirst for nutting out the picky details with the way I taught science in the early part of my career. In those days, I planned all classes myself, from scratch. I would have an idea for how to teach, say, gravity, give it a go, and it would or would not work. If students made mistakes or asked questions, I would sometimes respond effectively and sometimes not. It was all a bit hit-and-miss.
This is the culture of teaching. Countless teachers with four or five classes to plan for the next day face blank sheets of paper and just make something up. What will keep the students busy? Is this a good activity? To help address this burden, teachers Google worksheets that may or may not be well suited to their students, and websites such as Teachers Pay Teachers sprout up to monetize the exchange of resources.
As mentioned above, Direct Instruction is notable for the emphasis it places on teachers fully and clearly explaining, demonstrating, and modeling what they want students to do before asking them to do it. This stands in stark contrast to popular methods such as inquiry learning, where students are asked to investigate questions or problems and try to figure some of the concepts themselves. Some may draw a connection between the way teachers tend to plan classes and the popularity of inquiry learning—as a teacher, I know it is possible to keep students busy investigating, researching, and presenting their ideas, perhaps with a poster, without having to devote as much time and attention to the picky details of producing a more structured plan.
When Noel Pearson was looking for a way to improve outcomes in schools in Australia’s far North, he was impressed with what Direct Instruction had to offer. In Pearson’s account, “I came to see the evidence favored teachers actually teaching … it was the performance of the children that convinced me. I could see children learning once their teachers started teaching.”
Noel Pearson is an Indigenous Australian who trained as a lawyer and has worked extensively on the issue of Indigenous land rights. In 1990, Pearson co-founded the Cape York Land Council, and he provided legal advice to the Aboriginal and Torres Strait Islander Commission before turning his attention to the cause of education reform. Signaling his intent in an influential 2009 essay, Pearson rejected the idea that poverty is destiny and that before we can improve educational outcomes, we must first fix the issue of disadvantage. He wrote that:
It is too early to give up on the classical ideal that educational improvement can take place despite socio-economic disadvantage. We do not have to fix all social and economic problems to close the achievement gap. Indeed, the opposite is true: educational progress is an antecedent condition to overcoming broader social and economic disadvantage.”
Since then, Pearson has focused on improving outcomes for indigenous students in Cape York, a remote peninsula in Australia’s north, far from the main population centers—Pearson grew up at Hope Vale, a Lutheran mission in Cape York. In 2010, Pearson jointly founded Good to Great Schools Australia, which partners with the Cape York Aboriginal Australian Academy—a school with two campuses in Cape York—to deliver Direct Instruction programs as part of a package of reforms. Good to Great Schools Australia also coordinates a wider network of schools attempting to implement these reforms.
Whether or not Direct Instruction is the solution to education in remote parts of Australia with large Indigenous populations, reform of some kind is needed. Data from the 2018 round of the Programme for International Student Assessment (PISA) demonstrates that indigenous students in Australia are two to three years behind their non-Indigenous counterparts in reading, mathematics, and science.
Direct Instruction was one of the competing approaches pitted against each other in the largest education experiment ever conducted—the Project Follow Through study of early education that ran from the late 1960s to the early 1970s. Impressively, despite being categorized as a “basic skills” program, Direct Instruction not only had the largest positive impact on basic skills compared to control conditions; it also had the largest positive impact on more complex cognitive skills and self-esteem. Some of the programs in the experiment that were explicitly designed to improve cognitive skills had a negative impact on those skills and some explicitly designed to improve self-esteem had a negative impact on it.
It is not hard to understand why Direct Instruction would raise self-esteem in parallel with academic skills. It feels good to get better at something. Conversely, you can attempt to boost children’s self-esteem all you like, but if they are not making progress then this will start to ring hollow.
Nevertheless, Direct Instruction remains controversial in Australia. In an influential 2014 essay, Allan Luke, Emeritus Professor at Queensland University of Technology and an expert in “multiliteracies,” took aim at Direct Instruction. While begrudgingly acknowledging that Direct Instruction “can generate some performance gains in conventionally-measured basic skills of early literacy and numeracy,” he listed “many” criticisms. These criticisms range from “focusing on teacher control” to the “deskilling of teachers” and an opposition to tracking—Direct Instruction programs often use placement tests to determine which program is most suitable for each child.
Perhaps most tellingly, Luke criticized Direct Instruction because “it places the teacher and child in a rigid relationship where the teacher is always the one with the power and knowledge with limited allowance or recognition of individual and cultural difference.” For many education academics and bureaucrats, this is a deeply ideological issue. Direct Instruction offends the way they want education to be. Its effectiveness in raising literacy and numeracy levels for disadvantaged students is secondary.
For educationalists steeped in the enduring tradition of progressive education, Direct Instruction is offensive.
For these reasons, there is intense interest in the Australian education research community in the success or failure of Pearson’s Direct Instruction initiative. In 2020, two Australian academics released a paper purporting to demonstrate that Direct Instruction had not lifted literacy and numeracy levels when compared to similar schools that had not adopted the program and, alarmingly, that it appeared to be associated with poorer school attendance.
However, these findings were subsequently challenged in the same journal on the basis that the “post-intervention” data the authors had used was collected from the start of the intervention period.
Good to Great Schools Australia has also been the ongoing focus of press attention due to its funding arrangements that have seen Australia’s federal government support the program. Earlier this year, allegations of mismanagement were leveled at the organization, but they were not substantiated by an internal review.
The fact is that we will never establish to everyone’s satisfaction the exact results of Pearson’s initiative. It is not part of a randomized controlled trial but is, instead, an intervention in a messy, real-world system.
Extraordinary as it may seem, as a trainee teacher, I was never taught about Project Follow Through, and I am far from alone. One of the justifications I have heard for its omission from teacher education programs is that, although it was a controlled experiment, it was large and messy, with large variations in outcomes for the same program in different schools. We can always explain away the evidence, no matter how suggestive, if we are motivated enough to do so.
However, the effectiveness of Direct Instruction aligns with the wide body of evidence from educational psychology and from correlational studies of teacher effectiveness that favors explicit approaches where teachers fully explain concepts and model tasks before asking students to do the same. Direct Instruction is not the only way of teaching explicitly, but it is understandable that Pearson would turn to this solid bet when attempting to address the problems of disadvantage in remote Australian schools.
Academics in faculties of education will continue to focus on big, abstract ideas. They will continue to dislike effective programs because of oppression or something. I will continue to pay attention to those picky details.
This is a companion discussion topic for the original entry at https://quillette.com/2022/05/14/direct-instruction-works-so-why-is-it-controversial/