Development and testing of a new hostility scale

Evidence suggests that assessment of global type A behaviour as a coronary risk factor may not be the most fruitful strategy because some components, particularly hostility, may be more harmful than others. Measurement of hos tility has been primarily through two types of instrumentation: the Type A Structured Interview (SI) and the Cook-Medley (Ho) scale. Deficiencies in t he measurement of hostility have resulted from a lack of conceptual clarity because definitions of hostility have been blurred by overlap with the con cepts of anger and aggression. The main objective of this study was to deve lop and test an instrument that focuses hostility on attitude rather than e motion or behaviour. Psychometric properties of the Hostile Attitude Scale (HAS) were Cronbach's alpha, 0.80, and correlation of 0.56 with the Ho scal e and 0.43 with the SI. In a clinical investigation that used the HAS, SI a nd Ho scale, all measures were significantly correlated with angiogram outc ome, although the strongest relationship was with the HAS. Hostile attitude alone (HAS) accounted for 6% of explained variance in angiogram outcome ba sed on regression analysis. The HAS is offered as a short, easily administe red measure of hostility, with particular emphasis on cognitive aspects of the construct.