| This study investigated the relationships between patients' locus of control for health-related behaviors and three medical interview outcomes when menopausal women encountered directive or nondirective physicians. Using an analogue method, the study predicted differential outcomes for mood, information recall, and intended action. Subjects were a community sample of women aged between 41 and 59 years who were experiencing early signs of menopause, but were not yet postmenopausal.;The major findings were that internal expectancy and need were positively, and not, as hypothesized, negatively associated with dysphoria in the nondirective physician condition. In the directive condition, internal need was, again counter to prediction, negatively associated with dysphoria. Of the independent variables, only internal expectancy significantly predicted recall in the directive physician condition. Internal expectancy and need were positively associated with conformity in the nondirective physician condition. Internal need was negatively associated with conformity in both physician conditions. When physician directiveness was held constant, need was a better predictor than was expectancy of both dysphoria and conformity; whereas neither need nor expectancy significantly predicted recall. Finally, high need and low expectancy predicted negativity toward menopause.;To create the independent variable, expectancy, the Multidimensional Health Locus of Control Scale (Wallston et al. 1978) was adapted to make it more broadly applicable to expectancies of control of health-related behaviors. A parallel need scale was developed to measure Need Value and permit comparison of its predictive power relative to Expectancy. Control variables included age, education, hormone replacement therapy status, social desirability, attitudes toward menopause, symptom severity, and physician gender. Subjects were randomly assigned to listen to an audiotape of an interaction between a directive or nondirective, male or female physician before completing the dependent measures. |
