Clinical and quantitative classification of learning disabled children




Peter, Barbara Mary

Journal Title

Journal ISSN

Volume Title



It is now commonly acknowledged that learning disabled children are not a homogeneous population, and current neuropsychological research in this area has focussed on attempts to identify subtypes of these disorders. Earlier subtyping studies adopted a subjective, clinical-inferential approach to classification, while recent research has emphasised a strictly objective, quantitative model which involves the use of multivariate statistical methods of classification. In the natural sciences, both objective quantification as well as the developmental, qualitative aspects of taxonomy are considered equally essential for a good classification (Adams,1985). Subtypes identified so far in various studies have had relatively little impact on either neuropsychological theory or clinical practice, and this has been attributed to a failure on the part of researchers to integrate the clinical-qualitative approach with the quantitative subtyping procedures (Wilson & Risucci,1986). The present study attempted to address this problem by using a combination of these two generally accepted methods, in an attempt to identify reliable and meaningful subtypes within a sample of 275 clinic-referred and 26 normal control subjects. Two separate typologies of this subject sample were generated: a) using clinical-inferential methods, based on clinical inspection of psychometric test data, and b) using multivariate statistical methods for the derivation of subtypes (cluster analysis). The two subtyping solutions were then compared, allowing each to be used to validate the other (Morris & Satz,1984). The derived clusters and clinical typology groups identified all commonly found subtypes as well as most others reported by previous subtyping studies in the literature. The results of several internal validation procedures indicated that the clusters were relatively heterogeneous, and therefore somewhat unreliable, although the majority of clusters proved to be meaningful and interpretable. Comparison of the two classifications indicated approximately 58% correspondence in terms of individual case assignments to comparable subtypes between the typologies. Comparison of T-score ability profiles revealed generally satisfactory correspondence between the profiles of cluster analysis derived subtypes and those of comparable clinical subtypes. Further analyses were performed on selected groups of subjects in order to explore specific hypotheses. Age effects on subtype patterns were examined, and the results suggested that subtypes do persist over the school age range. However, adolescent subjects were more prominent in the severe language disorder subtypes, and a large proportion of the younger subjects emerged in subtypes characterized by visual perceptual problems. Reading disability subtypes were also analysed, indicating qualified support for Denckla's (1977) subtypes from the cluster analysis, but considerable confirmation of this typology from the clinical classification. Rourke and Finlayson's (1978) findings in regard to specific arithmetic disabilities were not replicated in this study. Subjects with specific profile patterns were also examined for evidence of characteristic social, emotional and behavioural difficulties, with mixed results. Finally, the obtained subtypes were examined in order to determine possible differences in terms of academic performance, in order to establish external validity for the two classifications. It was concluded that, although there are definable as well as meaningful subtypes of learning disabilities, this population of children cannot be classified into discrete subtypes with clear boundaries and strictly defined criteria. In addition, it was deemed important to recognize that such disorders range, in degree of severity, from quite subtle to seriously impaired, so that diagnostic "cut off" points are inappropriate for this particular group of children.



Learning disabled children