Analysis of large scale time use survey with respect to travel demand and regional aspects

Abstract

The time use study of the German Bureau of Statistics (Statistisches Bundesamt) is a remarkable data set. More than 19000 persons participated in the survey and recorded their activities for two consecutive days. The diaries have the form of a sequence of 288 activity codes, one for each interval of five minutes. The aim of the analysis is the modelling of travel demand. Therefore the connection between time use and geographic variables is of special interest. The study is based on a new developed cluster algorithm which was adapted to the special structure of the data set. So clusters of different size and heterogeneity can be identified. The clusters are characterised by the average time budgets and the sequence of activity blocks. In a second step socio-demographic groups are formed with respect to different activity participation. It is interesting which variables are the most meaningful to describe these groups. Furthermore, the peculiarities of different types of districts in Germany are presented. The spatial characteristics regarding the cluster analysis are discussed.