Experimental Intervention Research on Students with Learning Disabilities: A Meta-Analysis of Treatment Outcomes

Help with Listen Feature Help with Listen Feature

NICHCY’s Structured Abstract 36 describes the following:

Title | Experimental Intervention Research on Students with Learning Disabilities: A Meta-Analysis of Treatment Outcomes

Author | Swanson, H.L., & Hoskyn, M.

Source Review of Educational Research, 68(3), 277-322.

Year Published | 1998

This article summarizes a comprehensive synthesis of experimental intervention studies that have included students with learning disabilities. Effect sizes for 180 intervention studies were analyzed across instructional domains, sample characteristics, intervention parameters, methodological procedures, and article characteristics. The overall mean effect size of instructional intervention was positive and of high magnitude (M = 0.79). Effect sizes were more positive for a combined model that included components of direct and strategy instruction than for competing models. Interventions that included instructional components related to controlling task difficulty, small interactive groups, and directed responses and questioning of students were significant predictors of effect size, and interventions that varied from control conditions in terms of setting, teacher, and number of instructional steps yielded larger effect sizes than studies that failed to control for such variations. The results are supportive of the pervasive influence of cognitive strategy and direct instruction models for remediating the academic difficulties for children with learning disabilities.

In the two decades between the mid 1970s and the mid 1990s, the number of children classified as having learning disabilities (LD) nearly tripled. Students with learning disabilities became the largest single category of children served under special education, accounting for nearly half of all the children receiving special education services. Despite the growing number of children receiving special education under the category of “specific learning disability,” the authors of this study found that simple questions such as, “What are the components of effective interventions for students with learning disabilities?” or “Which intervention works best for which subgroup of students with learning disabilities?” could still not be answered. Some researchers and educators have supported the use of the “top-down” approach of strategy instruction*, while others have promoted the use of the “bottom-up” approach of direct instruction. This meta-analysis* was undertaken to explore these and other questions.

In this study the authors defined direct instruction as including at least 4 of the following components in a lesson:

  1. breaking down a task into small steps,
  2. administering probes,
  3. administering feedback repeatedly,
  4. providing pictorial or diagram presentations,
  5. allowing for independent practice and individually paced instruction,
  6. breaking instruction into simpler phases,
  7. instructing in a small group,
  8. modeling of skills by the teacher,
  9. providing set materials at a rapid pace,
  10. providing individual child instruction,
  11. teacher asking questions, and
  12. teacher presenting new (novel) materials. (pp. 283-284)

Strategy instruction was defined to include at least 3 of the following elements:

  1. elaborate explanations (i.e. systematic explanations, elaborations, and/or plans to direct task performance),
  2. modeling from teachers (verbal modeling, questioning, demonstration),
  3. reminders to use certain strategies or procedures (i.e. cues to use taught strategies, tactics, or procedures),
  4. step-by-step prompts or multiprocess instructions,
  5. dialogue (between teacher and student),
  6. questions from teachers, and
  7. provision by teacher of necessary assistance only. (p. 284)

The authors defined combined models as any instructional approach that integrated any 4 of the direct instruction elements with any 3 of the elements from strategy instruction.

Research Questions
The purpose of this meta-analysis was to identify effective interventions for students with LD. Specifically, the researchers sought to answer the following questions:

  1. Are treatments that include components of strategy instruction*, direct instruction*, or both methods more effective than other interventions?
  2. Are some instructional domains (e.g., reading) more resistant to intervention than others (e.g., social skills*)?
  3. Do studies that include samples with intelligence* and reading scores at various levels yield different outcomes than those studies in which such levels are not specified?

Research Design

  • Number of Studies Included | 180
  • Number of Subjects | Average sample size: 7.06 (SD = 40.15).
  • Years Spanned | 1972-1997

Research Subjects
The participants in these studies were students with learning disabilities.

Age/Grade of Subjects
Mean* treatment age = 11.6 years (SD = 3.22)

Specified Disability
Learning Disabilities (LD)

This meta-analysis investigates a variety of interventions for students with LD. Each intervention was classified into 1 of 4 models:

  1. strategy instruction and direct instruction
  2. direct instruction alone
  3. strategy instruction alone
  4. neither strategy instruction nor direct instruction

Duration of Intervention
The average duration across these studies:

  • 22.47 minutes of daily instruction (SD = 29.71)
  • 3.58 times a week (SD = 1.58)
  • for 35.72 sessions (SD = 21.72)


  • Experimental educational interventions in a variety of areas have positive effects on students with learning disabilities. However, the amount of change brought on by interventions is greater in some areas than others. The greatest levels of change found in this meta-analysis were in the domains of reading comprehension, vocabulary, and creativity.
  • Not all interventions were found to be equally effective. Combining the direct instruction and strategy instruction models was the most effective procedure for addressing the effects of learning disabilities.

Combined Effects Size
Across all studies the weighted mean effect size was 0.79 (SD = .052). Applying Cohen’s criteria for what constitutes a “marginal” effect size (an effect size below 0.45) , marginal effect sizes were found in spelling, mathematics, attitude, intelligence, social skills, perceptual processes, and language processes. Again applying Cohen’s criteria—this time for what constitutes a “large” effect size (an effect size over the threshold of 0.80)—only the domains of reading comprehension (0.72), vocabulary (0.78), and creativity (0.70) came close.

The results of this research suggest that students with learning disabilities benefit most from instructional approaches that combine elements of direct instruction with those of strategy instruction. This evidence suggests that the argument over a “top-down” versus a “bottom-up” model  fails to address the reality that elements of both approaches have merit and are most effective when used together.

According to the authors, the components that were most influential were:

  • Sequencing
  • Drill-repetition-practice
  • Controlling task difficulty
  • Segmenting information
  • Technology
  • Small interactive groups
  • Augmenting teacher instruction (for example, with homework)
  • Direct questioning/responding
  • Strategy cueing


* 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

NOTICE: The National Dissemination Center for Children with Disabilities (NICHCY) is no longer in operation. Our funding from the U.S. Department of Education’s Office of Special Education Programs (OSEP) ended on September 30, 2013. Our website and all its free resources will remain available until September 30, 2014.