Abstract

Current psychotherapy research on cultural humility (CH) indicates that CH is positively associated with client treatment outcomes and that working alliance may mediate the relationship between CH and treatment outcome. However, these studies have used cross-sectional data and have largely ignored the nested nature of therapist-client data sets. To address this gap in the literature, the current study applied a two-level, time-lagged, Multilevel Structural Equation Model (MSEM) to a longitudinal data set and examined whether working alliance mediates the relationship between client-perceived CH of the therapist and client-perceived symptom improvement (Outcome Rating Scale; ORS). The working alliance was conceptualized and operationalized as a dyadic construct (Cl-WA, Th-WA). Results were mixed; contrary to predictions, sessions in which clients perceived the therapist as higher in CH compared to average sessions were associated with poorer ORS scores. This relationship between cultural humility and client perceived improvement was not significant at the between-client level. The mediation hypothesis was supported at the within-client level but not at the between-client level. Implications for practice and research are discussed.