Moderators of congruent alliance between therapists and clients: A realistic accuracy model

Congruence between therapists' and their clients' alliance ratings was found to be beneficial to therapeutic processes and outcomes. To date, however, less is known about the possible moderators of such congruence. The current study adapted Funder's (1995) realistic accuracy model to identify a judge characteristic (therapists' affiliative tendencies), a target characteristic (clients' affiliative tendencies), information (time elapsed in therapy), and traits (bond vs. task/goal aspects of the alliance) that may moderate this congruence. These were examined using the innovative truth-and-bias model (West & Kenny, 2011), which allows the simultaneous estimation of two different congruence indices within repeatedly measured data: therapist/client temporal congruence (i.e., the correlation over time between therapists' and their clients' alliance ratings) and directional discrepancy (i.e., the average difference between therapists' and their clients' alliance ratings across sessions). Clients (n = 109) and therapists (n = 62) at a university-based clinic rated their affiliation tendencies at the beginning of treatment and rated their alliance perception after each session. Time elapsed in therapy, as well as therapists' (but not clients') affiliative tendencies were linked to higher therapist/client temporal congruence and to lower therapist directional discrepancy. In addition, congruence was higher for the bond aspect of the therapeutic alliance than for goals/tasks. Consistent with Funder's model, multiple factors (including judge, information, and trait) were associated with therapist/client congruence in alliance.