Lifestyles, and Risk Perception Consumer Behavior

Abstract

In this article, the concept of lifestyle is traced to its early roots in personality psychology and in marketing. In the latter field, many commercial marketing firms have made strong claims as to the explanatory power of lifestyle dimensions, often based on procedures which have been kept secret, but researchers have seldom been able to verify such claims. In spite of this, the approach is very popular, has wide credibility and is often given very favorable media coverage. Probably because of this, it is often considered as a very important and promising approach by administrators working with the regulation of risk and risk communication. It may also be credible in some quarters because it affords a way of ‘explaining’ risk perception as being non-rational. In this paper, we give results from an empirical study of nuclear waste risk perception which is related to a basic risk perception model and three approaches to lifestyles: Kahle's List of Values, a Swedish adaptation of the ‘Agoramétrie’ approach suggested by a group of French researchers, and Dake and Wildavsky's Cultural Theory dimensions. It was found that nuclear waste risk perception could be modeled successfully with risk attitudes and perception data (basic model about 65% of the variance explained), but that lifestyle dimensions added virtually nothing to the explanatory power of the model. Lifestyle dimensions in isolation only explained a minor part of the variance.

This study was supported by a grant from the Swedish Council for Planning and Coordination of Research and by the Bank of Sweden Tercentary Fund.

Notes

1. Sjöberg & Montgomery (1999) showed that the apparently good fit of expectancy models of attitude is misleading and that beliefs and values do not have the static character assumed in such models.

2. There is a trend towards more ambitious sampling. However, the field is still struggling with methodological problems. In a comparison of Japanese and US data (Hinman, 1993), a response rate of some 35% was accepted for the Japanese data. Response rates in the range 20–30% have been accepted in other work. See the methodological discussion by Sjöberg and Drottz-Sjöberg (2001).

3. McGuire (1985) has pointed out that advertising industry which was worth about 50 billions dollars per year hardly could document anything else than very marginal effects on attitudes and consumer behavior.

4. Yet, there is probably wide-spread belief in the marketing profession and the public that psychodynamic concepts and theory can be used to devise very effective advertising, e.g., by using more or less covert sexual themes.

5. A dimensional analysis is given by Vinson et al. (1976).

6. This is an AIO (Activities Interests Opinions) approach, differentiated from value systems according to Rokeach or Kahle's (1983) LOV (List of Values), a modernized and shortened version of Rokeach's list. CCA refused to show their questions to Valette-Florence who could only work with the classification which they presented. Commercial reasons apparently dictate much secrecy among those who market systems of segmentation based on life styles. Kahle (1991) points out that this is true also of SRI's (Stanford Research Institute) life style methodology, VALS II, which is being kept secret. The COFREMCA/CCA system is a commercial system which is common in France and is being exported to various European applications in the form of ‘Euro-Socio-Styles’. Some Swedish advertising agencies seem to be skeptical about life styles but when some of them are bought by foreign firms they may be changing some of their routines and start to use life style systems developed internationally. Media attention is often more or less certain when it comes to new lifestyle approaches, thereby creating a market demand.

7. Even with such a low level as a few percent of explained variance it can still be possible that there is a certain practical value of segmentation, cp. (Novak, 1992); on the other hand the theoretical gain for understanding social phenomena is very marginal.

8. Horn does not explain how they were able to obtain 80% response rate with a questionnaire using 1000 questions. The attempts by CCA to develop a European system led, according to Valette-Florence (1989), to a questionnaire of 300 pages and a data base including 7000 variables. One gets the impression that energy has replaced theoretical analysis in this work. It is not clear whether the questionnaire used by CCA should be responded to in its entirety by each respondent.

9. These items were made available to us by Demoskop.

10. Cultural Theory suggests that there are two more ‘cultures’: fatalists and hermits.

11. We are grateful to the late Karl Dake for providing us with these scales.

12. However, we did run all analyses with those items included as well. There were no important differences as compared to the results that are presented here.

13. The scores were computed as means of ratings of items which had been responded to.

14. There were some variations also in the rating scales themselves, but for the present purpose they were all converted to a common 0–5 scale before proceeding with the analyses reported here.

15. Great economic profit is reaped from these methods. Rorschach was recently exposed as a failure on several empirical grounds, yet at least one million people are tested worldwide every year (Wood et al., 2003).