Abstract
Two approaches to interpreting relationships among psychosocial risk factors for coronary heart disease are contrasted: the profile approach which focuses on predictive validity and confounding; and the theoretical approach which emphasises internal validity. These approaches are explored by applying them to a set of correlations found among Type A behaviour, hopelessness, anger management and social contact in a female sample. Type A behaviour was positively correlated with hopelessness and with anger expression, whilst anger management was related to the availability of social contact. The relationships between Type A, anger expression and hopelessness also appeared to be moderated by degree of social contact. These results are interpreted from the profile perspective and from that of Price's cognitive social learning theory. The relative merits of the two approaches are considered and a case is made for encouraging more theoretically based research on coronary-prone behaviour.