Please use this identifier to cite or link to this item:
Year of Publication: 
Series/Report no.: 
IÖB-Diskussionspapier No. 3/07
Westfälische Wilhelms-Universität Münster, Institut für Ökonomische Bildung (IÖB), Münster
The concept of the life styles has gained important meaning in many facets in social sciences in the course of the last century. Worked out at first on the basis of the hermeneutic tradition of the social sciences, it was backed up increasingly in the last three decades by empiric cluster analyses. In the course of its change this concept appears as a householding-style and especially as a housekeeping style in the householding sciences. By the latter, one understands the division of the households in types on the basis of the sequences as to the way time is used. The first empiric work was presented by Hufnagel in 2004 with data of the Time Survey Data of the statistical Federal Office Germany from 2001/2002. The present analysis is based on the Time Survey Data 1991/92 and limits itself to the single households. At first the distances between the day courses of the singles are defined on the basis of the provided diaries in which each of them has documented two days divided into a 5-minute-rhythm. Afterwards a cluster analysis by which the households are joined on the basis of resemblances in their day courses is carried out. Besides, one succeeds to a great extent in reproducing the results of Hufnagel (2004) who carried out a suitable analysis for the Time Survey Data 2001/02 and limited himself likewise to the singles. This contribution contains graphic and tabular representations of the found clusters and socioeconomic characteristic features. In addition, by means of a discrimination analysis applied on the clusters, the autonomous explanation contribution, which the cluster affiliation performs for the explicatory socioeconomic dimensions, can be proved. Furthermore that proves that the cluster-analytic approach raises the prediction strength of the socioeconomic theory compared with the exclusive use of more conventional multivariate analyses at least for the singles of the Time Survey Data 1991/92. Thereupon a far-reaching parallelism to Hufnagel's analysis can be shown.
Document Type: 
Working Paper

Files in This Item:
251.59 kB

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.