Abstract
The purpose of this study is to refine the theory of individuals' sense‐making of risk as a product of structure of meaning, and subsequently to test the theory empirically by identifying coherent groups of individuals by means of an operationalisation of each individual's structure of meaning. To this end, a two‐step cluster analysis was conducted on a random selection of the Swedish population (N = 778) aged between 16 and 75 that resulted in three groups of individuals: the ‘locally rooted’, the ‘globally minded’ and ‘the cosmopolitan’. A one‐way analysis of variance (ANOVA) was then used in a simple test of the viability of the operationalisation. The analysis shows that the three groups of individuals differ in their behavioural patterns in the face of various everyday risks. The article's principal contribution is the development of the theory of the structure of meaning's role in how individuals sense‐make risk. Furthermore, with the analysis and testing inherent in operationalisation, the study is an empirical contribution to the field. In future, the structure of meaning would bear closer study, and the operationalisation should be further refined to create a more sensitive gauge of sense‐making and its connection with the behavioural patterns associated with various risks.
Acknowledgements
The author wants to thank Anna Olofsson and Roine Johansson, Risk and Crisis Research Center at Mid Sweden University, for comments on earlier drafts. Also, Rolf Lidskog, Mikael Klintman and the anonymous reviewers were highly constructive in their feedback.
Notes
1. This response percentage should be seen as normal for a Swedish mail poll, bearing in mind that the questionnaire in question was quite lengthy.
2. The crises that the respondents had to consider were: fire at home; natural disaster; violence or aggression; war or terrorism; traffic accident; cancer, coronary or other life‐threatening disease; and accident as a result of a leisure activity.
3. In the solution deemed best, each variable results in high loadings for only one factor, while the solution as a whole is simple and logical. It was chosen despite the fact that only two variables are available for three of the four factors. Normally an optimal factor solution would have at least three variables per factor.
4. The total so divided numbered 643 respondents. The discrepancy arises because the chosen method for cluster analysis does not include responses with information loss. This means that any respondents who have not answered one or more of the questions used in the cluster analysis are excluded from the analysis in its entirety. Hence, in the present study, 144 respondents were removed from the cluster analysis of a total population of 778 respondents.