Searching for the best cognitive model for instructing students with learning disabilities: A component and composite analysis

Abstract

The purpose of this paper is to identify the components of various cognitive instructional models that best predicted effect sizes for students with learning disabilities. Based on an exploratory analysis, eight models were identified: Direct instruction (sequencing & segmentation), explicit strategy training, monitoring, individualised remedial training, small interactive group instruction, teacher-indirect instruction, verbal questioning/attribution instruction, and technology mediated instruction. The models were tested in a hierarchical weighted least square regression analysis. The results indicated that explicit strategy instruction (explicit practice, elaboration, strategy cuing) and small group interactive settings best predicted the magnitude of treatment outcomes. Although direct instruction also positively predicted the magnitude of effect size, its effects were mitigated when an explicit strategy model was entered into the equation. The results support the robust effects of strategy training as a general heuristic for instructing students with learning disabilities.