IF THE MODEL FITS ‐‐ AN ANALYSIS OF THE STRUCTURE OF THE REVISED ILLINOIS TEST OF PSYCHOLINGUISTIC ABILITIES FOR MODERATELY MENTALLY RETARDED CHILDREN?

A factorial study of the revised Illinois Test of Psycholinguistic Abilities was carried out with 98 moderately mentally retarded school children to determine the domain mapped by the subtests and whether these fit the theoretical psycholinguistic model of channels, processes and levels of communication. Raw scores from the 10 subtests together with Binet IQ were subjected to three methods of analysis: the alpha factor analysis, image analysis and principal component analysis to obtain “method independent” and more meaningful results. Factor loadings from the alpha factor and image analyses were tested for congruence by the Schonemann method and a factor “reliability” study was carried out with separate principal component analyses on two random subsamples of 49 children each. Results support the channel separation postulated by the theoretical model. Implications for language programming for the moderately mentally retarded are discussed. The need for a taxonomic study of psycholinguistic abilities of retarded and non‐retarded children within the framework of current views of developmental psycholinguistics is stressed.