Statistical methods for exploratory multidimensional data analysis on time use

Abstract

After a brief introduction on the different approaches between exploratory multidimensional data analysis (EMDA) and confirmatory multivariate data analysis (CMDA) the paper deals with EMDA and in particular multi-way methods suitable for comparing statistical studies when each of them has many variables observed on many cases as in time use data analysis. Then the paper focuses on the two main time-use analyses: transversal and sequential analysis using respectively 3-way data analysis and time-points cumulative frequencies tables and graphs. Application concerns differences in time use of categories of partners based on Istat – Time Use Surveys data 1988-89 and 2002-2003.