Abstract

In surveys, more particular, value studies, the question asking for religious belief is considered as a standard demographic term. These questions are usually dealt with in a standardised way without many methodological concern. In this study, the question asking for religious belief was three times altered. The data originate from three independent samples of Flemish 20-21-year olds (spring 1991, 1992, 1994). These three samples are, based on the comparability of several background variables, comparable. Results of the analysis show that the question for religious belief is very susceptible to change in response alternatives. For instance: the number of Catholics halved after adding the category ‘believing’ to the response alternatives, this while the wording of the category ‘catholic’ was invariant across the three samples. When adding ‘active’ to the category ‘free-thinker’ this category was reduced to one third of the original size, in the same time the category ‘non-believers’ doubled. These findings call for more methodological attention for this kind of questions. Some explanations are proposed and some propositions are made to do further research on this and related topics.

How to Cite:

Waege, H. & Loosveldt, G., (1995) “Drie keer geloofsovertuiging: de bevraging bevraagd”, Tijdschrift voor Sociologie 16(2), 181–192. doi: https://doi.org/10.21825/sociologos.86220