The Analysis of the Structure of University Positions in Poland Using Classification Methods

Abstract

Categorical data analysis is a statistical method that can be successfully applied in different scientific areas, such as: social, medical, psychological and political sciences. Classification and segmentation are statistical methods that usually have been used for large quantitative datasets to identify segments in the data, however if applied for categorical data for contingency tables, one may arrive at impressive results as well. This paper presents the use of classification and segmentation methods for categorical data in a contingency table based on real data from Central Statistics on the number of university positions in Polish voivodeships. The authors compare the results of different approaches and provide graphical results using advanced visualization tools, perceptual map (biplot) and dendrogram. Comparative analysis provides information on corresponding categories of academic positions in different voivodeships. All calculations are conducted in R.(original abstract)