Direct instruction

Based on Wikipedia: Direct instruction

The Teaching Method That Won—and Nobody Wanted to Hear About It

In the late 1960s, the United States government ran what might be the largest educational experiment in history. Called Project Follow Through, it compared more than twenty different teaching approaches across impoverished communities nationwide. The goal was simple: figure out what actually works for disadvantaged kids.

One method won decisively. It outperformed every other approach on every measured outcome—reading, math, spelling, and even students' self-esteem. You might expect such a method would sweep the nation.

Instead, it became one of the most controversial approaches in education, criticized as robotic, authoritarian, and soul-crushing. Many teachers actively despise it.

This is the strange story of Direct Instruction.

What Direct Instruction Actually Is

Let's start with some terminology that trips people up. When educators write "direct instruction" in lowercase, they mean any explicit teaching—lectures, demonstrations, tutorials. When they capitalize it as "Direct Instruction" or abbreviate it as DI, they're referring to a specific, highly structured approach developed by two men named Siegfried Engelmann and Wesley Becker in the 1960s.

The difference matters enormously.

Traditional direct instruction is just a teacher explaining things clearly. Direct Instruction with capital letters is something far more radical: a complete system where every word the teacher says is scripted in advance, lessons are choreographed down to the minute, and students respond in unison to rapid-fire questions.

Picture a classroom where the teacher holds a script and says exactly what it tells her to say. She snaps her fingers or claps to signal students to respond together. "What word?" Snap. "Cat!" the class responds in chorus. "Spell it." Snap. "C-A-T!" The pace is brisk, almost frantic. There's no improvisation, no tangents, no waiting for the class to settle down.

It looks, to many observers, like the opposite of everything progressive education stands for.

The I-We-You Framework

At its core, Direct Instruction follows a deceptively simple three-part structure that educators sometimes call "I do, We do, You do."

First, the instructor demonstrates. She works through a problem or reads a passage while students watch. This is the "I do" phase—pure modeling.

Then comes guided practice. The teacher and students work through similar problems together, with the teacher providing immediate feedback and correction. This is "We do"—scaffolded support.

Finally, students practice independently while the teacher monitors and intervenes only when necessary. This is "You do"—the gradual release of responsibility.

None of this sounds controversial. Teachers have been gradually releasing responsibility to students for centuries. What makes Direct Instruction distinctive is how precisely and systematically it implements this cycle.

The Birth of DISTAR

The story begins at the University of Illinois in the mid-1960s. Siegfried Engelmann wasn't a trained educator—he was an advertising executive turned preschool teacher who became obsessed with how to teach disadvantaged children more effectively. Carl Bereiter was a psychologist. Jean Osborn rounded out the team.

They started a preschool program for children from impoverished backgrounds. The sessions were short—just twenty to thirty minutes daily—and focused intensively on language, reading, and math. The approach was relentlessly explicit. Nothing was left to chance or discovery.

The children improved dramatically. Not modestly. Not incrementally. Dramatically.

Encouraged, Engelmann and Becker formalized their approach into a complete curriculum they called DISTAR—Direct Instruction System for Teaching Arithmetic and Reading. Every lesson was scripted. Every response was anticipated. Every correction was specified.

The name tells you something about the era. DISTAR sounds like a Cold War military project, all efficiency and systems thinking. And in a way, that's exactly what it was: an engineering approach to a problem that most people thought required art and intuition.

Project Follow Through: The Experiment Nobody Wanted to Believe

When the federal government launched Project Follow Through in the late 1960s, it was supposed to be an extension of Head Start—providing continued support to disadvantaged children as they entered elementary school. But budget cuts transformed it into something else: a massive comparative study of different educational philosophies.

The experiment ran for years across multiple sites. DISTAR was implemented in nineteen different locations, ranging across different demographics and geographic settings. The competition included progressive approaches emphasizing discovery learning, child-centered methods, and open classrooms.

When the results came in, they weren't even close.

Direct Instruction was the only intervention that showed significantly positive impacts on all measured outcomes. Not just academics—self-esteem too. The supposedly dehumanizing, creativity-crushing method produced students who felt better about themselves than those in child-centered programs.

The education establishment largely ignored the findings.

Why the Results Threatened the Status Quo

To understand the backlash, you have to understand what Direct Instruction implies about teaching. The method's foundational premise is radical: if students aren't learning, the problem is the instruction, not the students. All children can learn. All teachers can teach effectively—if given proper training and materials.

This sounds inspirational, but it has uncomfortable corollaries.

If teaching can be scripted and systematized, what does that say about teaching as a profession? If following a script produces better results than teacher creativity and judgment, what's the value of teacher expertise? If any trained instructor can achieve the same results, what justifies paying teachers professional salaries?

Critics called it "teacher-proof" curriculum, and they didn't mean it as a compliment. The phrase suggested that Direct Instruction treated teachers as interchangeable parts in a machine—warm bodies to deliver pre-designed lessons. Many teachers experienced the scripts as demeaning, a vote of no confidence in their professional judgment.

The Criticism That Wouldn't Go Away

Jonathan Kozol, the celebrated education writer, attacked a related program called Success for All in his book The Shame of the Nation. He called it excessively dogmatic, utilitarian, and authoritarian. Success for All wasn't pure Direct Instruction, but it shared the scripted approach—ninety minutes of reading instruction where teachers followed a pre-ordained lesson plan minute by minute.

The criticism touched a nerve because it connected to deeper anxieties about education. Should schools prepare students for standardized tests or nurture their creativity? Should teaching be a science or an art? Should we trust data or intuition?

Teachers in urban schools raised concerns about cultural sensitivity. The scripted lessons, developed largely by white researchers, didn't always resonate with diverse student populations. Critics asked whether efficiency in teaching basic skills was worth sacrificing relevance and connection.

What the Numbers Actually Show

Despite the controversy, researchers kept studying Direct Instruction, and the results kept coming back positive.

A 1996 meta-analysis—a study that synthesizes results across many individual studies—found an average effect size above 0.75, which statisticians consider substantial. A 2018 analysis by Stockard and colleagues found effects around 0.6 standard deviations, meaning Direct Instruction students performed significantly better than control groups.

To put that in context: an effect size of 0.6 means that a typical Direct Instruction student outperforms about 73 percent of students in traditional classrooms. That's not a marginal improvement.

John Hattie's landmark 2009 synthesis, Visible Learning, summarized four meta-analyses covering 304 studies of more than 42,000 students. The average effect size was 0.59—larger than any other curriculum approach Hattie examined.

The method works for nearly every population studied. In Project Follow Through, it ranked first for poor students and non-poor students, urban and rural students, African American, Hispanic, and Native American students. Today, many of the highest-performing schools run by the Bureau of Indian Affairs use Direct Instruction materials.

The Special Education Connection

One area where Direct Instruction found relatively welcoming ground was special education. For students with learning disabilities, the explicit, systematic approach addressed many of the challenges that make traditional instruction difficult.

Studies of single-subject designs—research focused on individual students rather than group averages—found substantial effects for students with learning disabilities. The structured repetition, immediate feedback, and carefully sequenced progression aligned well with how many of these students learn best.

Interestingly, research also found that students with higher IQ scores sometimes did better with a different approach called Strategy Instruction, which teaches students explicit problem-solving strategies. For students with lower IQ scores, a combination of Direct Instruction and Strategy Instruction produced the best results.

This suggests something that ideological debates often miss: different approaches work better for different students in different contexts. The question isn't whether Direct Instruction is "good" or "bad" but when and for whom it's most effective.

The Practical Objections

Beyond philosophy, Direct Instruction faces practical barriers. The materials are expensive. Student workbooks run about twenty dollars, and teacher materials can cost between $180 and $232. For schools in low-income areas—precisely the populations the method was designed to help—these costs can be prohibitive.

There's a painful irony here. A method proven most effective for disadvantaged students is often too expensive for disadvantaged schools to implement.

Training is another challenge. Direct Instruction requires teachers to master specific techniques: the pacing, the signals, the corrections. Done well, it looks effortless. Done poorly, it's a disaster—mechanical and confusing. Schools need ongoing support and coaching to implement it effectively.

In Australia, where Direct Instruction has been used in remote Indigenous communities, critics pointed out another problem: the materials' American-centric content felt alien to Aboriginal cultures. Teaching techniques that work in Baltimore may need significant adaptation for the outback.

What Direct Instruction Isn't Good For

Even proponents acknowledge limits. Anne Tweed, former president of the National Science Teachers Association, questioned whether Direct Instruction was the most effective approach for science education. Her point: direct instruction alone cannot replace the depth of understanding that comes from inquiry-based exploration.

This aligns with a broader principle in cognitive science. Different types of learning require different approaches. Memorizing basic math facts is different from understanding mathematical reasoning. Decoding words is different from comprehending complex texts. Learning to solve well-defined problems is different from tackling open-ended challenges.

Direct Instruction excels at teaching explicit skills with clear procedures. It's less suited for developing creativity, critical thinking, or the ability to navigate ambiguity. A complete education probably requires multiple approaches.

The Ninety-Ten Principle

One underappreciated aspect of Direct Instruction is its approach to practice and review. In a typical lesson, only ten percent of material is new. The remaining ninety percent reviews what students have already learned.

This sounds inefficient. Why spend so much time on review?

The answer lies in how memory works. New information is fragile. Without repeated retrieval and application, it fades. The constant cycling back through previously taught material builds what psychologists call "overlearning"—practicing skills well beyond the point of initial mastery until they become automatic.

When basic skills are automatic, they don't consume cognitive resources. A student who has to consciously think about sounding out each word has little mental capacity left for comprehension. A student for whom decoding is automatic can focus entirely on meaning.

Direct Instruction's relentless review isn't inefficiency. It's strategic investment in automaticity.

The Grouping Controversy

Direct Instruction groups students by skill level, determined by assessments administered before the program begins. A third-grader reading at first-grade level works with other students at that level, regardless of age.

This practice—sometimes called ability grouping or tracking—is controversial in its own right. Critics argue it can become self-fulfilling prophecy, with lower groups receiving lower expectations and less challenging instruction. There are also concerns about equity: ability groups often correlate with race and socioeconomic status.

Proponents counter that teaching students at their actual level, rather than their grade level, is essential for learning. A struggling reader placed in grade-level instruction may fall further behind while the class moves on. In a properly leveled group, that same student can master foundational skills before advancing.

The debate reflects a genuine tension. Heterogeneous grouping promotes equity and social integration. Homogeneous grouping enables targeted instruction. Direct Instruction firmly chooses the latter.

English Language Learners and the Spanish Bridge

Direct Instruction developed specific programs for English Language Learners, particularly Spanish-speaking students. The approach begins with instruction in Spanish, then gradually incorporates more English into the lessons—a technique called transitional bilingual education.

The theory is that students can learn reading skills in their native language and transfer those skills to English once they have sufficient vocabulary. Learning to decode Spanish words builds phonemic awareness that applies to English reading. Learning comprehension strategies in Spanish prepares students to comprehend English texts.

As with all Direct Instruction programs, the groups are kept small and organized by skill level rather than age. A newcomer with strong Spanish literacy might advance rapidly through the English transition. A student with limited schooling in their home country might need more foundational work.

The Baltimore Experiment

Some of the most compelling evidence for Direct Instruction comes from the Baltimore Curriculum Project, a large-scale implementation in Baltimore's public schools from 1997 to 2003.

The schools served predominantly African American populations, with more than 75 percent of students qualifying for free or reduced-price lunch—a standard proxy for poverty. Many schools started in the bottom twentieth percentile for academic achievement.

Over several years of implementation, some schools improved from the sixteenth percentile to above the ninetieth percentile. Not all schools showed such dramatic gains, but the overall trajectory demonstrated that sustained improvement was possible even in the most challenging environments.

Critics note that such implementations involve many changes beyond curriculum—new leadership, additional resources, focused attention. Isolating the specific contribution of Direct Instruction is methodologically challenging. But the results at least demonstrate that the approach can work at scale in real schools, not just controlled experiments.

The Uncomfortable Truth

Here's what makes Direct Instruction so frustrating to discuss: the evidence strongly suggests it works, yet many thoughtful educators find it philosophically objectionable.

This isn't cognitive dissonance. It's a genuine values conflict.

You can believe that Direct Instruction produces measurable gains on standardized assessments while also believing that standardized assessments don't capture everything important about education. You can acknowledge its effectiveness for teaching basic skills while questioning whether basic skills should dominate the curriculum. You can recognize its benefits for disadvantaged students while worrying about its implications for teaching as a profession.

The strongest argument for Direct Instruction is practical: kids who can't read can't do much else in school. Mastering basic skills opens doors. If Direct Instruction helps struggling students master those skills faster and more reliably, that's enormously valuable—even if it's not the only thing education should accomplish.

The strongest argument against it is also practical: education involves more than measurable skills. Curiosity, creativity, collaboration, and critical thinking matter too. An education system optimized entirely for Direct Instruction might produce students who can decode text but don't love reading, who can solve math problems but don't think mathematically.

A Way Forward?

Perhaps the most reasonable position is also the most boring: Direct Instruction is a powerful tool that works well for certain purposes in certain contexts. It's not a complete philosophy of education. It's not appropriate for every subject or every student. But for teaching foundational skills to students who are behind, its track record is hard to argue with.

The tragedy is that ideological debates have often prevented students from benefiting. While educators argued about philosophy, generations of struggling readers remained struggling readers.

Maybe that's the real lesson of Direct Instruction: in education, as in medicine, we should pay attention to what the evidence shows works—even when it conflicts with our intuitions about how things should work.

The children in those preschool classrooms at the University of Illinois didn't care whether their instruction was progressive or traditional, creative or scripted. They just learned to read.

That has to count for something.