Developments in Survey Methodology Anuška Ferligoj and Andrej Mrvar (Editors) Metodološki zvezki, 15, Ljubljana: FDV, 2000 Validation by Means of Method Variation: A CFA-Approach to the SSA-Value-Typology of Schwartz Hans Waege 1 , Jaak Billiet 2 , and Stefaan Pleysier 3 Abstract General values belong to the traditional topics of survey research. Since the beginning of the nineties, the work of Schwartz is becoming one of the main standards in this field. In order to construct his value typology, Schwartz uses Smallest Space Analysis. However, this technique does not allow testing theories, whereas Schwartz does formulate a theory that can be tested in the framework of a CFA-model. In this study, we have used a dataset of members of the Christian Workers Movement in Flanders (N = 2,070). It is shown that these data do not fit the theoretical value typology of Schwartz. Both the different types of values and the dimensions cannot be found in the data. The article concludes with a taxonomy of values and suggestions for the operationalisation of values in surveys. 4 1 Introduction Since the work of Rokeach on the nature of human values (1973), values belong to the traditional topics of survey research. They have been enjoying great popularity, especially in social psychology. At the beginning of the 1980s, a second important step was taken. Hofstede published his study of values at an aggregated level. He makes a study of values over 53 nations or regions and presents a second order 1 Faculty of Political and Social Sciences. University of Gent, Universiteitstraat 4, B-9000 Gent, Belgium; Hans.Waege@rug.ac.be 2 Faculty of Political and Social Sciences. Catholic University of Leuven. E. Van Evenstraat 2b, B-3000 Leuven, Belgium. 3 Faculty of Political and Social Sciences. Catholic University of Leuven, Belgium. 4 This project is supported by the Scientific Research Fund Flanders (FWO) G.012598. and by the Flemish Community PBO98.