Visualizations and Compositional Data Analysis for the European Time Use

Abstract

Time use data is a specific data that is of interest to many researchers such as sociologists, psychologists, statisticians and etc. It is multivariate by nature and can be seen as a subtype of multivariate data called compositional data. Thus, time use data can also be analyzed and visualized by compositional data analysis techniques. This is rarely mentioned in literature; usually analyses and visualizations of time use data is shown in multivariate real space. However, this may lead to false interpretation of the data and even some multivariate statistical analysis cannot be directly applied before a transformation is made. The main contribution of this study is to show how an analysis might be used for time use data and what benefits might be gained from this new setting of the data set as compositional parts. Indeed, the results show that there some are differences between traditional techniques and compositional data analysis.