Date of Award


Document Type


Degree Name

Master of Arts (MA)




There has been a recent proliferation of outcorie research on cognitive-behavioral treatment packages; however, little attention has been directed toward investigating the active ingredients of such conglomerate packages. The present study was designed to investigate the relative contributions of verbal and imaginal cognitive mediators in reducing publicspeaking anxiety with other treatment factors held constant.

Therapists' manuals were constructed for the two tynes of intervention so that they were identical except for the focus on either verbal or imaginal mediation. The treatments drev/ from techniques suggested by Meichenbaum (1974a), Goldfried (1971), and Suinn (1975a; 1975b), but the parallel format of the interventions was novel. Both treatment conditions included relaxation training, within-session practice via imagined scenes, and betweensessions practice via assignments.

Thirty-five speech-anxious college students (17 volunteers obtained through advertisement and 18 prescreened recruits obtained from psychology courses) were pretested in a speaking situation with self-report and behavioral measures. Twenty-four of the subjects were assigned to four treatment groups of 6 members each with volunteer/ recruit status counterbalanced across groups. The remaining 11 subjects were designated as the delayed-treatment control group. Two of the treatment groups were assigned to the self-instruction (verbal mediation) condition and 2 groups were assigned to the imagery (imaginal mediation) condition. Tv/o pairs of co-therapists each led one treatment group in each condition for 5 weekly, hour-long sessions.

Results from the post-intervention assessment were analyzed by an analysis of variance with covariant adjustment for pretest levels of the four dependent measures. The analysis indicated that the interventions were neither differentially effective nor significantly more effective than no treatment according to conventional statistical significance levels. The treatment effect for self-reported speech-anxiety did approach significance. Investigation of pretest and posttest correlations among deoendent measures suggested a closer alignment of self-reported and behavioral anxiety-responding following treatment. In the analysis of control subjects' data following delayed treatment, there were significant treatment effects on two dependent measures.

Several difficulties in the current study were discussed, most noteably, brevity of treatment and a less-than-adequate control group due to assignment constraints. Suggestions for future research in cognitive-behavior modification were made, including the need for further component analyses.