Causal attributions for heart disease: A systematic review

Abstract

The aims of this systematic literature review are to describe the pattern of attributions made for the causes of heart disease, and to determine how this pattern varies with the method by which attributions are elicited, and the respondent group.

A search yielded 47 papers and reports, containing 54 datasets. Lifestyle factors and chronic stress were the most common causes cited across all datasets. Attributions to stressors and fate or luck were more likely to be reported in studies that used interval rating scales than in studies that used dichotomous ratings. Cardiac patients were more likely to mention stressors and fate or luck as causes of heart disease; non-patients rated being overweight and hypertensive as more important

The differences observed between the responses of patients and non-patients may be due to actor - observer differences, or to a methodological difference: patients are often asked to report their own experiences whereas non-patients are asked about the general case.     

Consistent Treatment of Variables and Causation Poses a Challenge for Behavioral Research Methods: A Commentary on Nesselroade and Molenaar (2016)

Nesselroade and Molenaar presented the ideographic filter as a proposal for analyzing lawful regularities in behavioral research. The proposal highlights an inconsistency that poses a challenge for behavioral research more generally. One can distinguish a broadly Humean approach from a broadly non-Humean approach as they relate to variables and to causation. Nesselroade and Molenaar rejected a Humean approach to latent variables that characterizes them as nothing more than summaries of their manifest indicators. By contrast, they tacitly accepted a Humean approach to causes characterized as nothing more than summaries of their manifest causal effects. A non-Humean treatment of variables coupled with a Humean treatment of causation creates a theoretical tension within their proposal. For example, one can interpret the same model elements as simultaneously representing both variables and causes. Future refinement of the ideographic filter proposal to address this tension could follow any of a number of strategies.