The effect of meta-attitude manipulation on attitude change in psychotherapy



Journal Title

Journal ISSN

Volume Title



The present study was designed to investigate the usefulness of applying a particular model of attitude change, the Levy-House (1970) two-process model of attitude change, to the problem of effecting attitude change within psychotherapy. The Levy-House model implies that a patient's attitude about himself may become more susceptible to change if the attitudes he has about the validity of his self-attitude, i.e., metaattitudes about the self-attitude, are manipulated towards a belief that such self-attitudes are of questionable validity. The Ss were 24 psychiatric inpatients at the Veterans Administration Hospital, Houston, Texas. They were selected on the basis of questionnaire responses indicating a self-attitude of worthlessness or of being a failure, i.e., an attitude of negative self-esteem. Screened out as potential Ss were patients whose records indicated more than three hospitalizations for psychiatric purposes, patients whose primary diagnosis was drug or alcohol addiction, patients who were suspected of having brain damage, and patients over the age of 55. Ss were divided at the median of meta-attitude (MA) scores and within each level of MA they were paired on the closest match of selfattitude scores and randomly distributed to experimental and control treatment groups. No differences among the resulting four groups could be found on other organismic variables. There were three measures of self-attitude: selected items from the Personal Beliefs Inventory (Hartman, 1968), the 'O-scale' from the 16 PF (Cattell & Eber, 1962), and the Rotter Level of Aspiration Board task (Rotter, 1954, pp. 129, 318). Measures of two meta-attitudes were constructed: a measure of the extent to which the self-attitude was perceived as originating in learning versus heredity and a measure of expectancy to change the self-attitude. Six psychotherapists, all of whom were on the Veterans Administration Hospital staff, administered the treatments. The first part of the experimental treatment consisted of instructions during two 'therapy' sessions that the self-attitude originated in learning. The second part consisted of two sessions of structured role playing designed to build positive self- esteem. The first part of the control treatment consisted of the therapist's listening to the S's description of his self-attitude for two sessions. The second part was patterned after the corresponding intervention for the experimental treatment. It was predicted that: (a) the two measures of MA correlate significantly; (b) experimental Ss increase MA score after the first part of the treatment; (c) self-attitude is unchanged after the first part of the treatment; (d) self-attitudes of experimental Ss become more positive in self-esteem after the second part of the treatment; and (e) experimental Ss who originally scored high on MA measure make the greatest changes in self-attitudes in a positive direction. Hypotheses (a), (b), and (e) were not supported. Hypothesis (c) was supported, and Hypothesis (d) received partial support. The results generally did not support the Levy-House two-process theory of attitude change. Implications of the results were discussed in terms of the limitations imposed upon the design of the present study by the requirements of the applied setting and in terms of other theories of attitude change. Although highly tentative, the present results appeared most supportive of Rokeach's (1968a) theoretical formulation. Implications for further research and for psychotherapeutic practice were also discussed.



Attitude change, Psychotherapy