A discussion of the limitations of the psychometric and cultural theory approaches to risk perception

Risk perception has traditionally been conceived as a cognitive phenomenon, basically a question of information processing. The very term perception suggests that information processing is involved and of crucial importance. Kahneman and Tversky suggested that the use of 'heuristics' in the intuitive estimation of probabilities accounts for biased probability perception, hence claiming to explain risk perception as well. The psychometric approach of Slovic et al, a further step in in the cognitive tradition, conceives of perceived risk as a function of general properties of a hazard. However, the psychometric approach is shown here to explain only about 20% of the variance of perceived risk, even less of risk acceptability. Its claim to explanatory power is based on a statistical illusion: mean values were investigated and accounted for, across hazards. A currently popular alternative to the psychometric tradition, Cultural Theory, is even less successful and explains only about 5% of the variance of perceived risk. The claims of this approach were also based on a statistical illusion: 'significant' results were reported and interpreted as being of substantial importance. The present paper presents a new approach: attitude to the risk generating technology, general sensitivity to risks and specific risk explained well over 60% of the variance of perceived risk of nuclear waste, in a study of extensive data from a representative sample of the Swedish population. The attitude component functioning as an explanatory factor of perceived risk, rather than as a consequence of perceived risk, suggests strongly that perceived risk is something other than cognition. Implications for risk communication are discussed. (author)