Abstract: | ![]() Suggested modifications of the theory of reasoned action, which conceptualize the functional operation of the attitudinal and normative components in intention formation as separate, separate but contingent, and inseparable, are discussed. The model structures are examined with regard to the underlying psychological assumptions (psychological meaningfulness) and with regard to the required scale levels for the model variables, so that the results of model tests are invariant under the admissible scale transformations (formal meaningfulness). Using a procedure that combines “optimal scaling” with hierarchical regression analysis, the model variants are tested for women's (n= 89) and men's (n= 89) intentions to use contraceptive methods (pill, intrauterine device, condom, and natural methods) that require the model variables to be measured on interval scales. The results strongly favored the theory of reasoned action above the other model variants. Only for the subjective norm model was a better model fit found for a belief-only model compared to the original product-sum model. Gender differences were found with regard to the salient belief sets and the explanatory power of the normative model component. |