Statistical approaches to the analysis of belief patterns

Behavioural economics investigates economic behaviour and the ways to influence it. Lay beliefs and understanding are important determinants of expectations and behaviour and shape acceptance of public policy, and their study is therefore an important module of the field. Surveys constitute the main approach to collecting information. The means used to elicit beliefs fall into several categories, including multiple-choice questions, self-reports about the degree of support for statements, free associations to ‘inducing’ concepts and judgement of proximity between concepts. The techniques used to analyse the resulting data include conceptual core identification, hierarchical clustering, multidimensional scaling and CART tree building. We illustrate each of these with studies regarding topics such as the meaning of inflation, the good-begets-good heuristic in macroeconomics, the perceived causes of the financial crisis and the nature of capitalism.