Multivariate Analyses of Clinical and Statistical Groupings

Abstract

The similarity between clinical and statistical methods for the grouping of reading and writing characteristics of children was discussed earlier in conjunction with the description of Dia-Gjessing and its factor analysis. There are also similarities between clinical and statistical methods when classifying people. Dyslexics can be grouped on the basis of practical clinical experiences and theory according to certain patterns of behavior and test profiles. Various statistical methods also group people by similarities in data profiles. The two methods used in the Bergen Project were Q-factor analysis, which employs correlational techniques, and cluster analysis, which utilizes euclidian distances. Only the results of the Q-factor analysis will be described here.