The relationship between pre-therapy adult attachment style and the quality of early working alliance in counselling

• Main Author: Smaga, Maria
• Published: University of East London 2011
• Subjects: 150
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.536626

The present study explored the associations between the adult attachment style of the client, and client ratings of the early working alliance in counselling, and whether characteristics of attachment style (i.e., dominant and grouped attachment style and degree of insecurity) were able to predict clients' appraisal of the early working alliance. A total of 31 counselling clients aged between 24 and 57 (M=37.50, S!D=9.50) were interviewed to determine their attachment style via the Attachment Style Interview (AST). Participants provided a rating of their perceptions of the therapeutic alliance after the third counselling session via the Working Alliance Inventory (WAI). Five hypotheses and two further exploratory questions were examined. Securely attached participants had significantly higher ratings of the early working alliance than those with insecure attachment style, who rated the working alliance progressively lower the greater their degree of insecurity. Correlation analyses that included variables from the WAI and ASI showed significant positive correlations between "secure" attachment style and WAI subscales, and negative correlations between the "anxious" grouped attachment style and the WAI subscales. Likewise, "angry-dismissive" attachment dominant style was negatively correlated with WAI subscales. Participants' "ability to make and maintain relationships" was negatively correlated with all subscales of the WAI. Of the attachment attitudes, "desire for company" was negatively correlated, and "constrains to closeness" and "mistrust" were positively correlated with WAI global. The predictive utility of attachment style groupings and characteristics was examined using stepwise regression analysis, with WAI global scores as the dependent variable. Knowing whether or not an individual was securely or insecurely attached (grouped security type) was able to predict 19% of the variance in WAI scores. If a participant was insecure, knowing whether or not they were "angry-dismissive" was able to account for a further 12% of the variance. The results show ASI would appear to be a useful instrument of adult attachment style that is recommended for future research and clinical practice. Limitations and opportunities for further research are discussed, specifically the need for purposive sampling to explore differences between attachment styles in more depth, and the need to account for therapist competence and treatment framework as potential moderating variables.