Evidence converges from a wide range of contexts to suggest that the causal beliefs held by an individual, or predominant in a group of individuals, are of considerable practical significance and theoretical interest. A good system for analysing naturally occurring expressions of causal belief would provide the possibility of testing and exploring this supposition without having to rely on possibly artificial questionnaire responses. A rationale for such a system is described and a study of the reliability of the coding procedure is reported. It has proved possible to formulate definitions of the five major dimensions identified in the attribution literature so that, while they are not all orthogonal, each provides unique information. The coding system is shown to have adequate reliability, and to generate substantial quantities of data. The findings from these data are generally in accordance with those obtained using more traditional methodologies, and provide substantive information about the subject population. The method is proposed as an effective technique for empirical investigations of the forms, stability and significance of naturally expressed causal beliefs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.