Empirical Study of the Circumplex Model Using the Clinical Rating Scale

Abstract

　This paper investigated some of the major hypotheses of David Olson’s Circumplex Model of marital and family systems. The sample was ten Japanese clinical families who received systems-oriented family therapy at Yodoyabashi Psychotherapy Center in downtown Osaka. The family members' interactions were directly observed before and after treatment. The Clinical Rating Scale (Olson and Killorin, 1985) was used as an instrument to assess major dimensions of the Circumplex model, namely adaptability, cohesion and communication.

　The results of this study indicated as follows. First, two independent judges achieved very high interrater correlation (r=0.94 on average). This suggests that adaptability, cohesion and communication dimensions are clearly operationalized and highly communicative to even those judges who reside outside of the North American cultural context.

　Second, all adaptability variables, three out of four cohesion variables and six out of nine communication variables showed high item-total and inter-item correlations. However, three communication variables (freedom of experience, speaking for self and for others) were found uncorrelated to the total communication score. Thus their membership to the communication dimension was empirically questioned. Similarly, although the Circumplex model predicted a high Parent-Child coalition characteristic of an "enmeshed" family system, the coalition was found rather closely associated with a "disengaged" family system. Theoretical and clinical considerations were presented to support this empirical finding.

　Third, related to the discussion of parent-child coalition, it was found that sum of the understandings of parent-child or marital subsystems are not equal to that of total family system.

　Fourth, item-other total correlation analysis as well as factor analysis supported that the cohesion and adaptability dimensions were orthogonal (independent) of each other.

　Fifth, communication variables were found to be curvilinearly associated with adaptability and cohesion variables.

　Sixth, repeated-measurements of ten clinic couples before and after family therapy supported the hypothesis that the general health/ pathology of family systems is curvilinearly related to adaptability and cohesion dimensions.