A note on measurement issues in type A research: let's not throw out the baby with the bath water.

It has been suggested that researchers rely exclusively on the structured interview (SI) to assess Type A behavior instead of using objective self-report measures, because the SI is the only prospectively validated instrument currently available. This article considers the costs and benefits of relying solely on the SI to measure Type A behavior. Although using only the SI would assure that researchers measure, in a relatively unobtrusive fashion, actual Type A behaviors known to be predictive of heart disease, it would dramatically increase research costs, impede longitudinal studies of changes in Type A behavior, reduce the validity of statistical conclusions, restrict the convergent and discriminant validity of the Type A construct, and ultimately inhibit our ability to improve accuracy in predicting heart disease. A set of recommendations is proposed for improving the quality of measurement in Type A research.