Empirically Derived Learning Disability Subtypes: A Replication Attempt and Longitudinal Patterns Over 15 Years

Abstract

Test scores of two groups of learning disabled children with a mean age of 10 years (N=63 and N=96) were submitted to cluster analysis in an attempt to replicate previously described subtypes. Part of these subjects and a matched control group (N = 170) followed up into adulthood and tested at a mean age of 24 years were cluster analyzed to investigate the persistence of subtypes into adulthood. The clusters found in both childhood and adult samples did resemble subtypes reported in the literature, although at adult age the ‘linguistic’ subtype was no longer present. The results of several validation procedures indicated that the derived subtypes were somewhat less stable than desired. Tests for age differences across clusters shawed significant effects only with unstandardized variables in the adult sample but not in children. Tests for differences between derived clusters in neurological impairment as assessed in childhood proved significant only between control and LD clusters; a similar analysis with adult neurological categories showed signifcant trends related to severity. Tracing individual subjects from childhood to adult clusters showed only a moderate degree of persistence. A subsidiary analysis with two of the samples investigated the presence and outcome of the ACID subtype. One clear example of this type was found at both ages, but tracing subjects showed only little persistence of this type into adult years. A second example of the ACID type was also found at adult age which consisted mainly of subjects who were average learners in school. The paper stresses the need for reexamination of the notion of “specific” disabilities and further study of the replicability and long-term outcome of LD subtypes