Retraction of An et al. (2022)

Reports the retraction of "A strong alliance is not enough: Item-level variation in an alliance measure moderates the alliance strength and client outcome relationship" by Mira An, Dennis M. Kivlighan Jr. and Clara E. Hill (Journal of Counseling Psychology, Advanced Online Publication, Aug 08, 2022, np). The following article (https://doi .org/10.1037/cou0000629) is being retracted. This retraction is at the request of coauthors Kivlighan and Hill after the results of an investigation by the University of Maryland Institutional Review Board (IRB). The IRB found that the study included data from between one and four therapy clients of the Maryland Psychotherapy Clinic and Research Laboratory (MPCRL) who either had not been asked to provide consent or had withdrawn consent for their data to be included in the research. An was not responsible for obtaining and verifying participant consent but agreed to the retraction of this article. (The following abstract of the original article appeared in record 2022-87410-001.) This study investigated within-client effects of session-to-session working alliance (WA) strength (mean of client's and therapists' ratings of Working Alliance Inventory [WAI] items for a session; WAI-M) and intra-individual variance of working alliance (WAI-IIV; variation in how the same individual responds to different items in the WAI for a session) of therapist and client on overall client functioning. Specifically, we explored how the strength and intra-individual variance for therapist and client working alliance at a previous session (Time t-1) would relate to overall client functioning at a current session (Time t). We also explored whether the effect of WA-M on overall client functioning would be different at different levels of WAI-IIV. The dynamic structural equation modeling (Asparouhov et al., 2018) was used to analyze longitudinal data from 4,489 sessions at a university clinic where 17 doctoral student therapists provided low-cost, open-ended, individual psychodynamic psychotherapy to 135 adult community clients. We found that client-rated WAI-M and WAI-IIV had positive within-client main effects on next-session client functioning when controlling for autoregressive effects. Findings on WAI-M by WAI-IIV interaction effects revealed that the relationship between WAI-M at a previous session and client functioning at a current session was significant only when WAI-IIV was low (i.e., high intra-individual consistency across WAI items). Therapists' WAI-M, WAI-IIV, and interaction of WAI-M and WAI-IIV did not predict the next session client functioning significantly. Limitations and implications of the present research are discussed. (PsycInfo Database Record (c) 2023 APA, all rights reserved).