Direct Instruction

NICHCY’s Structured Abstract 1 describes the following:

Title | Direct Instruction

Authors | Adams, G., & Carnine, D.

Source | In H.L. Swanson, K.R. Harris, & S. Graham (Eds.), Handbook of Learning Disabilities (pp. 403-416). New York, NY: Guilford Press.

Year Published | 2003

The 2003 Handbook of Learning Disabilities is an edited compendium of contributions from over 55 of the leading scholars involved in learning disabilities (LD) research. Contributing authors were charged with reviewing the major theoretical, methodological, and instructional advances that have occurred in the LD field over the last 20 years. This particular article on direct instruction appears as Chapter 24 in the 4th section of the book, “Formation of Instructional Models.”

Students with learning disabilities (LD) differ in terms of their individual strengths, weaknesses, and learning styles. Therefore, no one instructional model can be recommended for all of them. Nevertheless, there may be certain general principles for teaching students with LD that can be adapted to meet each student’s individual needs. Successful interventions include strategies based on these general principles. These strategies can be adapted for different students, different content areas, and different settings.

This meta-analysis examines the use of one such successful instructional method: Direct Instruction (DI).* Direct Instruction is one of the most thoroughly researched instructional methods. It has been studied in both general and special education settings. This meta-analysis specifically examines the use of research-based DI programs for students who have learning disabilities. DI lessons are brisk-paced and include teacher modeling, group and individual responding, and student practice. Because not all teachers who use elements of direct instruction in classrooms are following a validated-DI program, the authors of this synthesis state that only studies of field-tested DI curricula were included in the synthesis.

Research Questions

  1. Overall, how effective is Direct Instruction for teaching math and reading to students with LD?
  2. Does Direct Instruction’s effectiveness vary depending on the age/grade level of the students?
  3. Do effect size scores for Direct Instruction differ depending on whether a norm-referenced or criterion-referenced test is used as a measure of effectiveness?
  4. Does the effectiveness of Direct Instruction vary depending on whether student groups are randomly assigned or not (experimental design vs. quasi-experimental design)?
  5. Are interventions that last longer than 1 year more or less effective than shorter interventions?
  6. How do trained experimental teachers and regular classroom teachers compare in the results they achieve with Direct Instruction?
  7. How do studies that are monitored in a laboratory setting differ in effect size from studies performed without monitoring?
  8. How does the performance of students with LD who are taught using Direct Instruction in the USA differ from students in other countries?

Research Design

  • Number of Studies Included | 17
  • Number of Subjects | N/A
  • Years Spanned | 1977-1990

Research Subjects
All students involved in these studies had learning disabilities and were taught using a research-based Direct Instruction curriculum.

Age/Grade of Subjects

Specified Disability
Learning Disabilities (LD)

All students were taught reading or math using a research-based Direct Instruction curriculum.

Duration of Intervention
Each Direct Instruction lesson lasted an average of 35-45 minutes. Nine of the interventions lasted a year or less, while 8 lasted over a year.


  1. Direct Instruction (DI) was an effective instructional method for teaching both math and reading skills to students with learning disabilities. DI was as effective for teaching students with LD as it was for teaching students who did not have learning difficulties.
  2. DI was most effective for high school students and adults.
  3. Students whose progress was assessed using criterion-referenced tests scored higher than those assessed using norm-referenced tests.
  4. Effect sizes did not differ significantly between experimental design studies, where students were randomly assigned to groups, and quasi-experimental designed studies using intact groups.
  5. Interventions lasting a year or less had higher effect sizes than those lasting more than a year.
  6. Students appear to respond better to DI instruction delivered by their regular classroom teacher than by teachers trained by researchers specifically for a research study.
  7. Direct Instruction had positive effects on student performance in controlled laboratory conditions as well as in real-world classroom settings.
  8. Direct Instruction has been found to be highly effective for students with LD in the United States as well as in other countries.

Combined Effects Size*
Overall Effect Size = 0.93 (large effect)

The results of this meta-analysis confirm what researchers have been consistently revealing for decades, that Direct Instruction (DI) programs are among the most effective research-based methods of teaching reading and math to students with learning disabilities. The fact that Direct Instruction programs demonstrate such overwhelming results should not be surprising considering that:

  1. The elements that make up the larger curricula are themselves grounded in scientific research, and
  2. Individual curricula undergo extensive testing before they are published for use by education professionals.

The authors suggest that Direct Instruction programs will provide the maximum benefit to schools that:

  1. are flexible in the way they organize instruction, allowing students to move between groups based on the rate at which they progress, and
  2. provide an infrastructure of training and support to ensure fidelity of implementation.

It is also worth noting that despite a strong research base, Direct Instruction is often met with resistance from education professionals who fear that using such a rigid approach would detract from the art and craft of teaching. The authors acknowledge this concern but are careful to warn against the adoption of alternative approaches without first evaluating the evidence to support its use.


* Terms Defined

Direct Instruction (DI) | A systematic, scripted form of instruction emphasizing lessons which are fast paced, sequenced, and focused.

Effect Size (ES or d) | A statistical calculation, often represented as ES or d, that measures the impact of an intervention. An effect size below d = 0.20 suggests that a treatment did not have a significant effect. An effect size of d = 0.20 is considered small or low; an effect size of d = 0.50 is considered moderate; an effect size of d = 0.80 or above is large.

Meta-Analysis | A widely-used research method in which (1) a systematic and reproducible search strategy is used to find as many studies as possible that address a given topic; (2) clear criterion are presented for inclusion/exclusion of individual studies into a larger analysis; and (3) results of included studies are statistically combined to determine an overall effect (effect size) of one variable on another.

Back to top