The hypothesis that separation anxiety is a prevalent affective undercurrent in the dreams and screen memories of panic disorder patients was supported. Significantly higher covert hostility in the panic disorder patients' dreams may support Bowlby's observation that people with high separation anxiety tend to disavow their anger.

In an attempt to test the hypothesis that therapist empathy is an important variable in successful dynamic therapy, the authors collected outcome measures and empathy ratings in the brief focal dynamic therapy of 59 patients. There was no significant agreement among patients, therapists, and clinical supervisors when they used the same scale to rate therapist empathy for the same sessions. Only the patients' ratings correlated significantly with some of the outcome measures, and they added modest but statistically significant predictive variance on multiple regression analysis. The authors discuss the implications of these findings for the traditional mode of supervision of dynamic therapy.

We adapted the Bem Sex-Role Inventory to survey therapists' beliefs about gender-appropriate behavior for hypothetical patients. We previously studied psychiatrists, finding that women were more likely to choose masculine or androgynous (high masculine and high feminine) traits as optimal for female patients while men were more likely to choose the undifferentiated (low masculine and low feminine) category for patients of either sex. In this study we used the same measure to compare psychologists and social workers to the psychiatrists, to determine the effect of medical education on these attitudes. We found that regardless of educational background, women were more likely to choose the androgynous category and men were more likely to choose the undifferentiated category. Women psychiatrists were more likely to choose masculine traits as optimal than were women non-physician therapists. Male therapists of all backgrounds were least likely to choose masculine traits as optimal for either male or female patients. Medical education per se does not seem to determine attitudes about gender-appropriate behavior. Men entering professions in which they are concerned about others' emotional well-being may have less stereotypically masculine beliefs about gender-appropriate behavior than women entering the same fields. Possible reasons for this difference are discussed.

Psychiatrists' beliefs regarding gender-appropriate behavior may influence their treatment of patients. Psychiatrists of both sexes (men: N = 76; women: N = 57) were asked to characterize optimal mental health for hypothetical female and male patients on the Bem Sex Role Inventory. The subjects' ratings for men and women were similar with two exceptions: more of the female psychiatrists rated masculine traits as optimal for female patients, and more male psychiatrists chose traits characteristic of Bem's undifferentiated category (low levels of both masculine and feminine traits) as optimal for both male and female patients. The results indicate significant changes in psychiatrists' attitudes toward gender in the past 20 years.