Profile analysis of a referral sample.

Abstract

This investigation demonstrates the use of Skinner's Modal Profile Analysis technique to uncover the underlying classification structure in a group of children (n = 235) referred for psychological evaluations by their public school teachers. Data were acquired on children's intellectual development, academic achievement, and social adjustment. The profile solution identifies three profiles of children with each profile containing two distinct types. Thus, six types (groups) were identified. This three-profile (six-group) solution is able to classify 81% of the referral sample. However, the statistical solution to the classification question did not correspond to the system by which the children were classified in their respective schools. While a discriminant analysis using profile elevation, scatter, and shape scores as the predictor variables and the child's classification in the school as the criterion of interest indicates a statistically significant ability to predict the child's school label, for all practical purposes the predictive ability is not useful.