Compared alprazolam, imipramine, and placebo in the treatment of panic disorder patients (n=59) to investigate concerns about the internal validity of the double-blind design. Found that the great majority of patients and physicians were able to rate accurately whether active drug or placebo had been given and physicians could distinguish between…

Addresses strategies needed in using longitudinal data on psychopathology to test cause-and-effect relationships, including natural experiments, testing of competing hypotheses on mechanisms, study of reversal effects, multiple replications in different circumstances, use of designs to dissociate possible mechanisms, testing for dose-response…

Gotlib and Asarnow suggest that learned helplessness may not adequately model clinical depression. They reply to Rohsenow's criticism that Gotlib and Asarnow's conclusions are unwarranted because the null hypothesis is not demonstrated by a single study and because adequate diagnostic criteria are not presented. (Author/BEF)