Marianne Welmers

A Systemic Perspective on Alliances and their Relation to Outcome 127 CHAPTER 5 When comparing our study findings with previous studies on unbalanced alliances, it should be noted that studies differ largely in how the concept of unbalanced or split alliances is operationalized. Like in our study, several studies used a continuous measure, such as a difference score (e.g., Forsberg et al., 2014; Robbins et al., 2003) or a similarity index (Glebova et al., 2018) to reflect the degree of unbalance in alliance between two family members. Other studies compare groups by defining the concept of split alliances in terms of standard deviations. In these studies, alliances are referred to as ‘split’ when the strength of two alliances differ at least one standard deviation (e.g., Muñiz de la Peña et al., 2009). The commonality among these studies is that they all adopt a quantitative approach to examine unbalanced or split alliances. Considering the fact that results of these studies are ambiguous, and that some studies specifically indicate that differences between family members in their alliances with the therapist are the rule rather than the exception (e.g., Welmers – van de Poll et al., 2020), a valid question seems: when do unbalanced alliances become problematic in achieving favorable outcome? Perhaps this distinction is qualitative rather than quantitative, and thus, future research could benefit from a more qualitative observational investigation of how problematic differences between family members’ alliances differ from non-problematic differences. Our findings regarding the family’s shared sense of purpose showed that the family’s shared sense of purpose was moderately but not significantly associated with the decrease of youth internalizing behavior problems 18 months after treatment. The fact that the correlation in our sample failed to reach significance, could indicate that the importance of the family’s shared sense of purpose is negligible. However, we should consider that the small sample size resulted in limited statistical power, and that the direction of the association was in line with meta-analytic findings of Friedlander et al. (2018), indicating that in couple and family therapy the family’s shared purpose is predictive of outcome. Taken together, our finding suggests that a stronger shared purpose within the family at mid-treatment may have a clinical significance in reducing internalizing child behavior problems that warrants further exploration. Although not a focus of this study, it was interesting to find that youths’ self-reportedmid- treatment alliances with the therapist were strongly associated with positive treatment outcome in terms of total behavior problems. This indicates that building strong alliances with children and adolescents seems vital for family treatment success in terms of youth behavior problems. However, strong alliances with children and adolescents in the context of conjoint family treatment is complex, as was illustrated in a previous study on alliance discrepancies between family members in home-based family treatment. In this study, alliances with children and adolescents were significantly weaker as compared to alliances with parents (Welmers - van de Poll et al., 2020). Furthermore, therapists were