Patterns in journal citation data revealed by exploratory multivariate analysis

The paper discusses techniques to emphasize patterns in citation data and to study their dynamics. In this context, the STATIS and the STATIS dual methods are presented. The methods are a generalization of the principal component analysis from a dynamical point of view. STATIS and its dual are applied in order to illustrate the dynamic of "citing-cited patterns" by using citation data of sixteen major journals from the statistics field.