Research Note: Extending the application of the psychometric approach for assessing public perceptions of food risk: some methodological considerations

Abstract

The 'psychometric approach' to the study of public perceptions of hazards has largely been limited to describing and exploring patterns of perception using cross-sectional survey data. This article discusses a number of methodological issues that need to be considered prior to employing this approach in studies monitoring changes in perceptions, the impact of risk communications, differences between groups, and other potentially more informative applications. Drawing on data from a cohort-sequential UK survey, this article considers item presentation order, sample conditioning, reliability and validity in an investigation ofpublic perceptions of food-related hazards. The data suggest that item presentation order has a major effect on ratings- in particular ratings of hazard familiarity and there is evidence that this is in part due to anchoring effects. The findings suggest that the approach produces an acceptable level of test-retest reliability and validity and they replicate previous research in this area (e.g. Sparks and Shepherd, 1994). The data also reveal how public perceptions of the BSE (bovine spongiform encephalopathy) hazard changed during the course of the UK crisis on just one of the two principal judgemental dimensions with BSE being perceived as significantly more dreaded but no less familiar. The implications are discussed in the context of general hazard/risk perception research.