AN EMPIRICAL TEST OF THE MULTIPLE COMPONENT MODEL OF IMAGERY IN PSYCHOTHERAPY

Use of mental imagery in psychotherapy is a rapidly growing trend that nevertheless is hampered by a number of unresolved theoretical and empirical issues. Information processing research is reviewed to illustrate that images are complex arrays of cognitive and affective information. Traditional misconceptions regarding mental imagery are discussed, including the notion that images are mental photographs which act as functional equivalents for their real life referents. Several research issues are presented in the context of the individual differences hypothesis and intrapersonal factors which influence the characteristics of imagery. A multiple component model of imagery is developed which holds that psychotherapeutic images are underpinned by empirically distinguishable sets of interrelated imagery abilities. A distinction is developed between image structure and image function. Image structure refers to the format of images; function refers to those processing operations employed to interpret and appraise information presented in an image. The clinical efficacy of imagery is argued to be attributable to the fact that persons in therapy learn to emit effective coping responses despite the presence of negative emotional arousal. Two studies are reported which evaluated key hypotheses of the model. Study One involved administering a battery of experimental and previously validated imagery self report measures in three one hour testing sessions to 199 undergraduate college students. Subjects were also asked to complete an analogue clinical visualization task which involved clinically relevant scenes. Psychometric evaluation of the measures revealed that high order reproductive imagery, sensory modality imagery, imagery control and cognitive style dimensions were factorily distinguishable ability dimensions. Measures of higher order reproductive imagery were significant predictors of performance on the analogue task; sensory modality imagery scales did not emerge as effective predictors. Indications of sex differences in the structural organizations of imagery were observed with females showing a greater preference for a general ability factor (G) than their male counterparts. Study Two involved the administration of analogue Systematic Desensitization, Covert Modeling and Controlled Exposure treatments to 55 test anxious college subjects. All subjects completed a pre-treatment battery of self report imagery measures and provided in vivo ratings of imagery vividness, emotional arousal and degree of coping imagery while imagining therapeutic scenes. Treatments produced significant and comparable reductions in test anxiety relative to a waiting list control group. Treatments appeared to positively influence cognitive/semantic evaluations of a simulated test session, but failed to influence state anxiety responses. Analysis of the relationship between imagery abilities defined by the self report measures and eventual success in the therapies revealed little evidence for the mediating role of reproductive imagery in therapeutic outcome. Analysis of state ratings revealed a modest trend characterized by an increased importance of emotional arousal and coping imagery at the latter stages of treatment. Results are discussed as providing limited support for the multiple component approach. Limitations of the present studies and implications for future research are discussed