Comparison of the toy-play behavior of autistic, retarded, and normal children: a reanalysis.

Abstract

Tilton and Ottinger (1964) examined differences among autistic, retarded, and normal children by observing their behavior in a toy-play setting. The purpose of this study was to reanalyze these data using a multiple discriminant function analysis, which allowed consideration of all 10 categories of toy play and their intercorrelations within one analysis. Significant differences ( p < .001) were found among the 3 groups and between the possible pairs of groups (normal-autistic, normal-retarded, autistic-retarded). In addition, information about the statistical classification of individuals was available. In the four discriminant function analyses, the proportions of Ss statistically classified the same as their original psychiatric diagnosis were .96 of the normals, .83 of the autistics, and .89 of the retardates. The combinational category of toy play emerged as the most important variable in discriminating the groups in all four analyses. It was concluded that this observational technique combined with the multiple discriminant function analysis would have practical utility as a diagnostic and evaluative measurement instrument.