Cognitive structural models: The perception of risk and prevention in coronary heart disease

Abstract

This paper examines the relationship between causal models and the assessment of actions in a naturalistic domain. Two studies used a novel network diagram technique to elicit individual causal models of the factors affecting the risk of coronary heart disease. The same individuals also rated the effectiveness of different preventive actions. In their diagrams, individuals depicted the causal interrelationships between various factors and estimated the strength of the connecting paths. Total path strength accounted for two-thirds of the variance in ratings of preventive actions. A number of methodological issues were also explored, which supported the utility of the technique. Consensual models allowed a ready comparison between groups of subjects and showed the importance of distinguishing the presence of paths from the strengths of those paths. The studies support the value of the technique in analysing individual differences in the way people represent complex causal structures; various extensions and applications are also proposed.