Interviewees were selected from the local population in the regions around the wilderness areas according to the theoretical-sampling method (Strauss 1991) in order to identify as contrastive views as possible (Hunziker 2000). The aim was not to select positions that are representative in a quantitative sense, but rather to have the greatest possible differences between positions. Thus interviewees were chosen to ensure that as wide a range of reasons as possible for opinions on the spread of wilderness would be covered. This selection procedure meant that all thematically relevant positions in the sampling universe could be identified and then taken into account in the sampling procedure (Fig. 1).
Fig. 1. Theoretical-sampling according to Strauss (1991). The small dark circles represent the extreme positions in the sampling universe that have to be considered in the theoretical selection (figure adapted from Hunziker 2000).
The first step was to prepare a hypothetical sample that would be relevant to the study, based on the literature on related areas. For this we drew on studies of attitudes to the re-introduction of wild predators (Egli et al. 2001) and of society’s requirements for forest areas (BUWAL 1999). This research has identified some of the demographic characteristics that are closely associated with different attitudes to these phenomena. These characteristics, in this case the age of the interviewees and their professional relationship with nature, are therefore likely to be relevant to research on attitudes to the spread of wilderness (Table 1).