Psychosocial functioning of learning-disabled children: replicability of statistically derived subtypes.

The PIC scores of 132 learning-disabled children between the ages of 6 and 12 years were investigated using Q-factor analysis, four hierarchical-agglomerative clustering techniques, and one iterative partitioning clustering technique. Results revealed excellent correspondence between the subtypes derived by all grouping methods in terms of both misclassifications and mean PIC profile similarity of the subtypes across techniques. The mean PIC profile of one subtype indicated normal psychosocial adjustment; a second subtype exhibited evidence of significant internalized psychopathology; a third subtype had a mean PIC profile suggestive of externalized psychosocial maladjustment. These subtypes were virtually identical to three subtypes reported by Porter and Rourke (1985) in another study of learning-disabled children. The results indicate that learning-disabled children comprise a heterogeneous population in terms of psychosocial functioning and that subtypes of learning-disabled children with similar patterns of socioemotional adjustment can be recovered reliably from this population across samples and statistical grouping techniques.