Marianne Welmers

Chapter 6 142 Christoph et al., 2006; Kazdin & Nock, 2003; Zilcha-Mano et al., 2016). Addressing this issue of reversed causality requires the use of early treatment improvement measures and specific methodologies to investigate the reciprocal effect between alliance and problem improvement (e.g., by means of autoregressive cross laggedmodeling; Glebova et al., 2011; Zilcha-Mano et al., 2014). Findings of a recent meta-analysis on this emerging body of research on individual adult psychotherapy indicates that alliance and symptom change indeed affect one another over the course of treatment (Flückiger et al., 2020). The review furthermore indicates the temporal precedence of a stronger alliance on symptom improvement, providing stronger indications for a possible causal effect of alliance on treatment gain (Flückiger et al., 2020). However, in family treatment, systemic alliance aspects particularly raise questions on reversed causality: congruence in family members’ attitudes toward treatment and the therapist (reflected in balanced alliances) and a within-family collaboration on shared goals may par excellence be early family treatment gains rather than predictors of eventual treatment outcome. Further research investigating this premise is warranted to disentangle the alliance-improvement cycle in systemic family treatment. Another limitation concerning causality is that alliance was examined as an isolated process factor. In reality, a successful therapeutic process of course entailsmany factors likely to mutually affect one another and to reciprocally affect treatment outcome (Liber, 2020). Examples include expectancies of treatment (Sprenkle & Blow, 2004), therapists’ theoretic orientation or adherence to a specific model (Karver et al., 2008; Lange et al., 2017), family empowerment (Hoagwood, 2005), and the use of specific therapeutic techniques (Chen et al., 2020; Liber, 2020; Weinberger, 2014). The relatively small effect size of the meta-analysis in Chapter 2 underscores the importance of a more integrative approach in future research, investigating the reciprocal effect of alliance and other factors over the course of treatment. This kind of research helps answering the important question of what causes desirable family treatment outcomes, and adds to our knowledge on key factors that should be addressed to improve effectiveness of (home-based) family treatment. What Works for Whom? Diversity Considerations An important strength of this dissertation is that the research was conducted in a naturalistic clinically representative context. The use of observations as well as the focus on therapists’ contributions enhances the translation of research findings into recommendations for clinical practice. Nevertheless, this dissertation does not answer the question when and for whom specifically alliance building behaviors contribute to better alliances, and ultimately to better treatment outcomes. To answer this question, future research could benefit from studying interaction effects between client and