Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)




The paradoxical techniques of encouraging and scheduling symptoms have "been used by therapists from a variety of theoretical orientations including communication systems, psychoanalysis, behavior modification, and existentialism. The various techniques have in common the scheduling of problematic behavior tmder specific conditions to promote behavior change. The study examined the effects of paradoxical instructions on state anxiety in an interview ajia- logue setting, The experiment was undertaken to assess the immediate impact of paradoxical instructions in a controlled situation. Existing studies are primarily of the case report variety and are hopelessly confounded with uncontrolled factors such as adjunctive chemotherapy, psychotherapy, and non-specific therapy effects.

Previous investigators have suggested that "spontaneous” behavior with autonomic components such as anxiety, blushing, and falling asleep cannot be consciously controlled. It was hypothesized that a group given paradoxical anxiety instructions would exhibit less state anxiety than a group which received non-paradoxical anxiety instructions or no specific anxiety instructions. Individuals from the three instruction treatment groups were seen in a twenty minute semi-structured interview in which dependent measures were taken in the initial, middle, and last four minute segments. Prior to the experimental session, the subjects were required to complete the Spielberger trait anxiety inventory and the groups were matched for trait anxiety.

The general experimental design consisted of a 3*2x3 factorial (treatments x sex x time segments), with the time segment dimension a repeated measure. The dependent variable (state anxiety) was assessed by means of a multi-method measurement strategy which included self-report, expressive speech, and behavioral indicators of state anxiety. Immediately following the interview, subjects completed the Spielberger state anxiety inventory and during the interview their speech and behavior was continuously monitored. Anxiety level manifested in speech was assessed using the non~ah ratio (speech disturbance index) while anxiety manifested in overt behavior was assed with ratings on a behavior checklist. The number of interviewer utterances and questions were also analyzed as a check for possible interviewer bias.

Results failed to support the experimental hypothesis. There v/ere no significant differences in state anxiety level among the instruction treatment groups or between the sexes on any of the anxiety dependent measures. Correlational evidence suggested that perhaps insufficient levels of stress v/ere used in the interview. The control group which received no specific anxiety instructions had significantly higher ah ratios which v/ere interpreted as reflecting a higher degree of uncertainty resulting from the instructions. There were significant treatment x time segment interactions for interviewer activity measures and these were discussed with reference to subject influence on interviewer behavior as a function of the experimental instructions given. Methodological problems were discussed and suggestions were made for further research.