Direct Instruction: What the Research and Teachers Really Say

In a 2025 RAND survey of more than 1,700 elementary teachers, just 2 percent of those using Connecting Math Concepts, Direct Instruction's math program, rated it negatively, the lowest dissatisfaction rate among elementary math programs surveyed. Only 1 percent of teachers nationwide reported using it.

Direct Instruction is among the most studied programs in education, and its scripted lessons make it a flashpoint in teacher preparation. The teacher reports suggest satisfaction tracks training and coaching more closely than the program itself.

For certification candidates, that leaves two practical questions: what a school funds beyond the box of materials, and how it stands with local universities.

What Direct Instruction Is (And How It Differs From Explicit Instruction)

Capital-D, Capital-I: A Specific Set of Programs

When researchers write "Direct Instruction" with capital letters, they mean a suite of commercially published programs for teaching reading, math, language and other subjects. The programs grew out of work Siegfried "Zig" Engelmann and colleagues began in the 1960s. Connecting Math Concepts, the math program teachers rated in a 2025 RAND survey, is one example.

The programs share a recognizable design:

• Scripted lessons: Teachers follow written wording meant to achieve what became known as "faultless communication," so every student hears the same clear explanation.
• Fast pacing: Lessons move briskly, with frequent student responses rather than long stretches of teacher talk.
• Skill-level grouping: Students are placed by what they already know, not by grade or age alone.
• Mastery checks: Students demonstrate a skill before the program moves them forward.

Lowercase Explicit Instruction: A General Approach

Lowercase explicit (or direct) instruction is broader: any teacher-led lesson in which the teacher models a skill, guides students through practice and gives feedback before they work independently. Many teachers know it as "I do, we do, you do." There is no required script, no publisher and no prescribed program. A teacher can use explicit methods with a district curriculum, a homemade unit or a textbook.

Why the Distinction Matters

The two terms often get used interchangeably, and that blurs the evidence. Studies of the formal programs test a full package: the scripts, the sequencing, the grouping and usually the training that comes with them. Those results cannot be assumed to apply to any teacher-led lesson. The reverse holds too: research on general explicit teaching does not automatically validate, or condemn, a scripted commercial program. When you read a claim about "direct instruction," check which one the author means.

How a DI Lesson Sounds

A Direct Instruction classroom has a distinct rhythm. The teacher presents a short piece of content, signals, and the group answers in unison. If some students respond incorrectly, the teacher corrects the error on the spot, models the right answer and has the group respond again until everyone has it. Then the lesson moves to the next item. That call-and-response pattern lets the teacher hear within seconds who has the skill and who needs another pass. It is also the feature critics point to when they call the method "robotic."

What the Research Says: Project Follow Through to Recent Reviews

Is Direct Instruction actually backed by strong research? The evidence is large and mostly favorable, but uneven in quality, and the strongest claims need qualifiers.

Project Follow Through

The federally funded Project Follow Through enrolled more than 200,000 children in 178 communities and initially compared 22 instructional models in the early grades.¹ Direct Instruction outperformed the other models and was the only one with significantly positive results across all outcome measures, including the project's measures of self-esteem and confidence.¹ A later analysis put the overall Direct Instruction effect at 0.63, drawn from 998 effects across 20 studies.²

The design had real weaknesses. Sites and comparison groups were not randomly selected or assigned, attrition was substantial, and the Abt Associates evaluation report acknowledged these limits. Results also varied by location: one analysis found negative effects at 3 of 10 sites despite a positive national average.

The Stockard Meta-Analysis

Stockard and colleagues (2018) reviewed research from 1966 to 2016, covering 328 studies, 413 designs and roughly 4,000 effect estimates.³ The overall effect size was 0.60, a medium effect.³ Reading, mathematics, language and spelling all showed significant positive results. Effects faded little after instruction ended, and academic gains grew with more exposure to the program. The authors reported that publication type, methodology and sample characteristics were not systematically related to results. The one exception to the positive pattern was a set of analyses involving affective outcomes.

Strong Versus Weaker Evidence

Treat the findings in tiers:

• Stronger: Large comparison studies and quasi-experimental work by independent evaluators, including the Follow Through data and re-analyses of it.
• Weaker: Dissertations, developer-affiliated studies and descriptive reports. The sources reviewed do not give a risk-of-bias rating for every study in the meta-analysis, nor show that it excluded nonrandomized or unpublished work.

For that reason, the 0.60 figure should not be read as the product of randomized trials alone.

Limits to Keep in Mind

• Independent replications are few compared with the total volume of studies.
• Coverage is uneven across subjects and grade levels, with the most evidence in early reading and math.
• Fit for secondary schools remains unclear.
• Critics also argue that a tightly structured scope and sequence can constrain teacher and student creativity, a point the effect sizes cannot settle.

The fair summary: consistent positive findings across decades, with design limits that justify confidence in the direction of the effect more than in its exact size.

Direct Instruction Effect Sizes by Subject and Study Quality

An effect size (Cohen's d) shows how far the average Direct Instruction student scored above the comparison group, in standard deviation units; by traditional benchmarks, about 0.2 is small, 0.5 is moderate and 0.8 or higher is large. Figures from a pooled meta-analysis, a single review of selected comparisons and a study limited to students with disabilities answer different questions, so do not rank them against each other as if they came from the same design. The strongest evidence is the 2018 half-century meta-analysis, which pooled hundreds of studies with consistent positive results. Single-program estimates, special-population studies and analyses co-authored by the program's founder carry more caveats, and the table flags them.

Why Direct Instruction Divides Researchers and Teacher Prep Programs

The fight over Direct Instruction centers on scripted lessons, in which the program spells out what a teacher says, asks and does at each step.

Where the Scripts Came From

Siegfried "Zig" Engelmann and his colleagues began scripting lessons in the 1960s, aiming for what became known as "faultless communication," instruction so clear that students could not reasonably misread it. Elaine Bruner, one of the first teachers to work with Engelmann, said teachers told the developers, "Don't tell us what to do. Tell us what to say." The scripts began as a response to teachers, yet they now draw the sharpest criticism.

What Critics Argue

• Autonomy: Scripts limit creativity, responsiveness and a teacher's ability to adapt a lesson to the room.
• Robotic delivery: Fast choral responses and heavy practice are said to crowd out exploration and intrinsic motivation.
• Transfer: Some question whether skills learned in tightly structured lessons carry over to unfamiliar or open-ended tasks. No definitive published finding settles this.
• Engagement: The "drill and kill" charge assumes students tire of repetition.

How Supporters Respond

Supporters call the script a template, not a prohibition. Teachers can respond to students, add explanations, use their own words and give extra practice, and experienced teachers can build on the structure. APM Reports describes the game-show rhythm as designed to hold attention. Jade Livers, a fifth grade teacher at City Springs, says she chose an education career after trying Direct Instruction and that critics who call it "robotic" would be more open if they visited classrooms.

On results, defenders cite a 2018 meta-analysis of 328 studies from 1966 to 2016 that found benefits in reading, math, language and spelling, with an overall effect of about 0.60 standard deviations.¹ That figure is an average of achievement. It says nothing direct about transfer, engagement or autonomy. APM Reports adds a caution from Project Follow Through: Direct Instruction had a positive effect nationally, but one analysis found negative effects at 3 of 10 sites, and no program succeeded everywhere.

The Divide in Teacher Preparation

City Springs Elementary/Middle School in Baltimore has used Direct Instruction for more than two decades, according to principal Rhonda Richetta. APM Reports describes how a local university told her it would not send student teachers there because of the school's use of the programs. The reporting does not name the university or explain its reasoning, so treat it as one concrete case, not a measure of how common the stance is.

The vocabulary adds confusion. Lee University, for example, describes student teaching moving from orientation and observation to participation and direct instruction, which is a general teaching term and not the scripted programs.² For candidates, the practical point is that a school's curriculum choices can shape which placements and hiring relationships are open to you.

What Teachers Report: The Spring 2026 RAND Survey on Connecting Math Concepts

Two adoption strategies look similar on a purchase order: buy a scripted math program and leave implementation to each teacher, or buy the same program and pay for the training and coaching that make it usable. Teacher survey data leans toward the second path, though the evidence comes with limits.

The Favorable Numbers

As APM Reports summarized, RAND's American Instructional Resources Survey, fielded in spring 2025 through RAND's American Educator Panels, asked more than 1,700 elementary school teachers nationwide about the materials they use in reading, math, and science. Among teachers who used Connecting Math Concepts, Direct Instruction's elementary math program, about two-thirds (67 percent) reported positive or very positive views.¹ Just 2 percent viewed it slightly negatively, the smallest negative share among the elementary math programs RAND surveyed.¹ Connecting Math Concepts also ranked highest in net favorability. Both figures describe users of the program, not elementary math teachers as a whole.

The survey is a broad look at instructional resources, not a program evaluation. It gives a national read on how users of one DI math program describe their views, at a time when program-level teacher feedback is scarce.

Who Answered, and What That Means

The caveats matter as much as the headline. Only about 1 percent of elementary math teachers nationally reported using Connecting Math Concepts.¹ The favorable responses therefore come from a small, self-selected minority: teachers who chose the program, were assigned it, or continued using it. Their experience is real, but it is not a random sample of math teachers. The survey also did not ask about Direct Instruction reading programs, so these results cannot be extended to the full suite of DI materials.

RAND measured perceptions, not student learning. The net-favorability ranking excluded respondents who held only slight opinions, positive or negative, so it is not simply another way of stating the 67 percent positive figure. And a favorability survey cannot establish that the program raises achievement. It tells us how a particular group of users describes the program.

The Wider Evidence on Teacher Views

Outside this survey, teacher-view data on Direct Instruction is thin. APM Reports found that much of it comes from graduate school dissertations, and that teachers generally describe positive experiences, particularly when they receive initial training and ongoing coaching. Read alongside the RAND results, that pattern suggests support may shape satisfaction as much as the scripted lessons do. A program can earn strong reviews from its users and still remain a minority choice.

Training and Coaching: The Variable That Seems to Matter Most

A school can buy the program and trim the coaching, and on a budget spreadsheet that looks like prudence. The implementation literature suggests the reverse: reports of positive teacher experience cluster where teachers received solid initial training and ongoing coaching, not where materials were simply purchased, though the evidence is associational rather than a clean test of coaching alone.⁴

What the support looks like in practice

In a well-run implementation, support is concrete rather than inspirational. Teachers get placement testing so students start at the right lesson, regular observation with feedback, and fidelity checks that confirm scripts are delivered with the intended pacing and correction routines. The National Institute for Direct Instruction describes peer coaches selected in the first year from teachers with strong practice, then trained across two to three years in data analysis, classroom observation and problem solving.¹ Their coaching is school-based and deliberately non-evaluative, so a teacher can flag a lesson that went sideways without it landing in a file. Monitoring appears elsewhere in reading instruction too: a 2026 Iowa survey of K-12 reading practitioners, not specific to Direct Instruction, found 57% received external monitoring, every respondent had received fidelity training and more than 80% had coaching or professional development support, with direct observation the most common method.²

When the support disappears

The clearest signal comes from implementation-fidelity research. A three-study analysis found that administrative decisions disregarding recommended protocols for teacher training, preparation time and student scheduling were associated with lower fidelity, weaker mastery progress and lower achievement, most findings significant and large in effect.³ In one district study, more extensive support predicted higher fidelity, and higher fidelity predicted schools staying with the program.³ Fidelity translated into achievement only when the schedule allowed regular, sufficient instructional time, which makes the principal's calendar part of the intervention.

The autonomy tradeoff

Scripted lessons shift the job rather than shrink it. Planning time drops because the lesson is written; the cognitive work moves into delivery, error correction and progress monitoring. Some teachers find that freeing. Others feel the loss of authorship. Both reactions are fair. Ask in interviews how much of the day is scripted and who decides.

For candidates, the practical question is not whether you like scripts in the abstract but whether the school funds the training and coaching that makes them work.

Related Articles

Direct Instruction Pros and Cons: Where It Helps and Where It Struggles

Read this table as a set of trade-offs, not a verdict: the same features that make Direct Instruction effective for some learners are the ones critics say limit flexibility for others. The research behind each cell varies in scope and setting, so weigh it against your own students and school context. Direct Instruction tends to break down when teachers get only a brief introduction to the program, when no one coaches them through the first months of scripted delivery, or when students are placed in lessons that do not match their skill level. In those conditions, the pacing and mastery checks that drive the positive findings below are hard to deliver.

What a Direct Instruction Curriculum Costs: Materials, Training and Coaching

The box of materials is the cheaper half of the decision. What a school actually commits to when it adopts Direct Instruction is a multi-year training and coaching line item, the part that tends to disappear first when budgets tighten.

What the published materials prices show

McGraw Hill Education publishes most Direct Instruction programs, including Reading Mastery and Connecting Math Concepts.¹ NIFDI, the implementation institute, does not sell them directly; its materials worksheet runs on 2025 pricing and tells schools to verify current figures with McGraw Hill sales. No complete published U.S. 2026 price list covers every component, so the figures here come from catalogues of different years and currencies and should be read as reference points, not quotes.

• U.S., 2023-revised schools catalogue: a Reading Mastery textbook at $83.95, a workbook at $28.95, and a teacher materials package at $1,179.95.² The catalogue extract does not label the level for several amounts.
• United Kingdom, 2026 catalogue (GBP): a Reading Mastery storybook at £50.60 and workbook at £17.18; a student textbook at £55.64, workbook at £22.19, teacher materials at £200.00, a Level D workbook at £15.00, and a digital student subscription at £8.39.³ No package totals are listed.
• Australia/New Zealand, 2026 catalogue (AUD): student textbook A$68.95, workbook A$32.95, teacher materials A$712.95.⁴

One reported estimate attributed to an IES figure puts initial student materials at $100 to $150 per student and teacher presentation sets at $600 to $1,000 per grade level.⁵ The reporting does not identify the publication year or program edition, so treat it as a planning band.

Separating one-time costs from recurring ones

Teacher presentation sets and the teacher materials package are largely a one-time purchase. Consumable workbooks are repurchased every year, per student. Training and coaching recur indefinitely if a school wants to maintain fidelity. McGraw Hill's Australia/New Zealand training page offers introductory training at no cost, running one to two hours depending on need, alongside longer packages (roughly eight hours comprehensive, six hours refresher, six hours online). A 2026 article reports NIFDI remote coaching at $121 per hour, data review at $62.50 per hour, and leadership academy sessions at $2,750, estimating $50,000 to over $100,000 in year one for coaching and monitoring at a mid-sized school.⁵ NIFDI's own pages do not confirm those rates.

The gap, and the cliff at the end of a grant

The same article compares a widely used unscripted core program, CKLA, at roughly $35 to $50 per student for consumables, $1,500 to $2,500 per classroom for teacher kits, and $2,500 to $5,000 per day for whole-staff professional development. Those figures are article-reported rather than independently established. The practical risk is structural: when coaching is grant-funded and the grant ends, the support that teacher surveys associate with satisfaction ends with it, while the scripts remain.

How M.Ed. Coursework Treats Explicit vs. Constructivist Instruction

No central database tallies how curriculum and instruction M.Ed. programs balance explicit instruction against constructivist or inquiry-based methods. That gap is itself informative: the balance is a program-by-program question, and one you can investigate before you apply.

Read the catalog for method language

Pull the required-course list and course descriptions, then scan for recurring terms: "direct instruction," "explicit instruction," "inquiry-based," "constructivist," "methods," "pedagogy," "practicum." A catalog listing both "Models of Inquiry" and "Systematic and Explicit Instruction" tells you the program blends approaches, as most do. A catalog with no methods language often signals that pedagogy is treated as general rather than as named approaches. Email the coordinator and ask directly how the program handles explicit versus inquiry methods, request sample syllabi and required reading lists, and check how many supervised fieldwork hours each program requires. Reading lists are the fastest tell: direct instruction texts such as Carnine sit differently next to Dewey, Vygotsky, or Brooks and Brooks. Methods also surface in placements more often than in course titles.

What each outside source is good for

• NCTQ Teacher Prep Review (nctq.org): comparative program ratings and standards coverage. Its 2026 review centers on early reading, not a general explicit-versus-inquiry classification¹, so read it for literacy coverage rather than a verdict on methods courses.
• The program's website and catalog: required coursework, syllabi, fieldwork hours.
• State education departments: program approval and licensure requirements.
• CAEP and AACTE: accreditation and program-quality context.
• NCTM, NSTA, NCTE and ASCD: position statements on instructional methods.
• BLS Occupational Outlook Handbook: teaching roles and job outlook.

Compare, then ask better questions

Line programs up on the same criteria: methods coursework, supervised fieldwork, and whether any faculty member can discuss direct instruction by name. Because most programs blend both traditions, the useful question is not which camp a school belongs to but whether it prepares you to choose deliberately and adjust when a lesson does not land. Do not wait for an external ranking that sorts programs by method. None exists.

Checklist: What to Ask Degree Programs and Employers Before You Commit

Teacher reports suggest satisfaction with Direct Instruction depends heavily on the support around it. Use these questions on admissions calls and in job interviews to find out what that support actually looks like.

• Initial training

  Ask the school: "What training do new teachers get before teaching a Direct Instruction or scripted program, and who delivers it?"
• Long-term coaching funding

  Ask: "Is ongoing coaching budgeted beyond the first year, or does it end when the adoption grant or startup funds run out?"
• Balance of scripts and discussion

  Ask: "Where do scripted lessons fit in the day, and how much time is left for discussion, inquiry and projects?"
• Student-teaching placements

  Ask your degree program: "Do you place student teachers in schools that use Direct Instruction, and if not, why?" Some universities have declined to send candidates to these schools.
• Coverage of explicit and constructivist methods

  Ask: "How does your coursework treat explicit instruction and constructivist approaches? Do candidates practice both, or only study one critically?"
• How fidelity is evaluated

  Ask: "How do you judge whether a lesson is delivered with fidelity? Is that feedback used for coaching or for formal evaluation?"