Searching for the best model for instructing students with learning disabilities

Focus on Exceptional Children, Oct 2001 by Swanson, H Lee

Students with learning disabilities (LD) comprise a heterogeneous group so no general instructional model can be recommended for all of them. Thus, the title of this article might seem to be somewhat of a misnomer. Nevertheless, some common general principles for teaching students with LD can be assumed to exist. Effective interventions include some instructional components that capitalize on these principles. Although these principles often operate in different ways with different students, in different content areas, and in different settings, they can be used in designing effective remediation programs for LD students.

In this article, findings are summarized related to a comprehensive educational intervention research synthesis for students with learning disabilities (Swanson, Hoskyn, & Lee, 1999) that attempts to identify the principles underlying effective instruction. Although there have been several technical reports related to this synthesis (e.g., Swanson. 1999; Swanson & Hoskyn, 1998; Swanson & Sachse-Lee, 2000), the purpose of this article is to consider more directly the practical aspects of the findings. More specifically, the focus will be on the question: What instructional components or activities characterize highly effective intervention programs for students with LD?

On the surface, identifying key instructional components that improve LD students' performance might seem to be a relatively simple, although time-consuming, task. One could simply take all the studies published in refereed journals that yield positive outcomes for students with LD and then summarize the instructional principles that make up those studies. In fact, most syntheses that attempt to translate research to practice rely on such procedures. Such an approach is problematic, however, because the published literature is biased toward reporting positive treatment outcomes (e.g., Begg, 1994).

A quick perusal of the published literature reveals a plethora of published treatments (i.e., bona fide instructional models) reflecting a diversity of approaches that yield positive outcomes for students with LD. Such a state of affairs has been called a "Dodo bird" effect (e.g., Lubrosky, Singer, & Luborsky, 1975). Taken from a conclusion of the Dodo bird in Alice in Wonderland (Carroll, 1965/1962), the bird stated, "Everybody has won, and all must have prizes" (p. 412). Unquestionably, this effect is disconcerting when attempting to search for some general principles of effective instruction.

In the case of children with LD, however, some treatments can be assumed to be less effective than others, and therefore not all treatments deserve prizes. How, then, can one determine the best approach to wade through all these studies that yield positive outcomes? One approach allowing comparison of the treatments that yield positive outcomes for students with LD is to place studies on the same level playing field. This is done by equating studies by their methodological sophistication (to be discussed later). If this is not done, the race as depicted in Alice in Wonderland becomes haphazard:

[The competitors] were placed along the course, here and there. There was no "One, two, three and away," but they began running when they liked, and left off when they liked so that it was not easy to know when the race was over. (p. 45)

Thus, the assumption is made that if studies are put on an equal footing methodologically speaking with control for publication biases by including nonpublished studies in our synthesis, we can identify some common components that make for a generally effective instructional model. Before detailing attempts to sort through the literature to identify those components, however, we would do well to review what has been found in previous syntheses of the literature.

PREVIOUS SYNTHESES

Reviews of the instructional literature that have been influential in providing an understanding of treatment outcomes for students with LD use a procedure called metaanalysis (e.g., Kavale & Forness, 2000; Mastropieri, Scruggs, Bakken, & Whedon, 1996). Meta-analysis is a statistical reviewing technique that provides a quantitative summary of findings across an entire body of research (Cooper & Hedges, 1994). The results of individual studies are converted to a standardized metric or effect size. The scores then are aggregated across the sample of studies to yield an overall estimate of effect size. Particular attention is given to the magnitude of the effect size estimate. According to Cohen (1988), .80 is considered a large effect size estimate, .50 a moderate estimate, and .20 a small estimate.

There have been several excellent meta-analyses on instructional research in learning disabilities (e.g., Kavale & Forness, 2000, Mastropieri et al., 1996), but none to our knowledge has considered intervention research across a broad array of academic domains and/or controlled for variations in methodology. The only synthesis to my knowledge that provided an overall estimate of treatment effectiveness prior to the present synthesis was our own previous synthesis (Swanson, Carson & Lee, 1996). A collection was assembled of published group-designed studies (78) between 1967 and 1993, which focused on youth 6-18 years of age; and reported from a total of 324 effect sizes a mean effect size of .85 for treatment versus control conditions. Using Cohen's (1988) threshold of .80 for large effects, the earlier meta-analysis suggested that various instructional approaches have had a significant beneficial effect when used with children and adolescents who have learning disabilities.