Investigations of metric and categorial data courses

Author(s): Franke, T.

Journal/Book: Physikalische Medizin Rehabilitationsmedizin Kurortmedizin. 1994; 3/4: 94-96.

Abstract: The continuous collection of data concerning the variables in question is of advantage over the isolated measurement of parameters at the beginning and end of the observed therapy. Thus it is possible to assess the dynamics of changes, to appreciate trends, and to recognize faulty measurements.In addition to the modified boxplot format for the presentation of course measurements, the mathematical tools of multivariate trend analysis and cluster analysis are demonstrated. Cluster analysis, especially in connection with the extended possibilities of graphic display of results is more convenient than in commercial implementations. Multivariate trend analysis allows the statistical investigation of variation of any given variable in level, shape, or dependence an time or treatment in the number of patients and groups under Observation. In display categorial courses, an intuitive possibility is demonstrated.

Keyword(s): Explorative data analysis

Zurück | Weiter