A multilevel meta-analysis of client and therapist predictors for alliance quality: Absolute and relative...

Extensive research, including meta-analytic studies, has underscored the role of working alliance in influencing psychotherapy outcomes. However, far fewer studies and meta-analyses have delved into the predictors of working alliance quality. To synthesize the literature and highlight key predictors for alliance, the present study conducted a meta-analysis of predictors of subsequent working alliance. Drawing on 898 Pearson's r effects from 130 studies involving 12,449 clients and 2,840 therapists, a multilevel meta-analysis identified five major groups of factors that preceded alliance. Among client-related factors, client resource and readiness positively correlated with alliance, showing small and medium effect sizes, respectively (r = 0.18, 95% CI [0.12, 0.24], p < .001 and r = 0.34, 95% CI [0.25, 0.42], p < .001), while client distress negatively predicted alliance (small effect size, r = -0.14, 95% CI [-0.18, -0.11], p < .001). Regarding therapist-related factors, treatment execution showed a small effect size (r = 0.15, 95% CI [0.03, 0.28], p = .019), and rapport building contributed to alliance quality with a medium effect size (r = 0.39, 95% CI [0.31, 0.47], p < .001). The study revealed notable between-study heterogeneity and found several moderation effects. The meta-analytic effects remained robust across various sensitivity analyses. This study illuminates the significant roles of both clients and therapists in developing working alliance and discusses its implications, limitations, and future research directions. (PsycInfo Database Record (c) 2025 APA, all rights reserved).