ISSN 0439-755X
CN 11-1911/B

Acta Psychologica Sinica ›› 2015, Vol. 47 ›› Issue (10): 1279-1287.doi: 10.3724/SP.J.1041.2015.01279

Previous Articles     Next Articles

The Developmental Patterns of Working Alliance in Counseling: Relationships to Therapeutic Outcomes

ZHU Xu; HU Yue; JIANG Guangrong   

  1. (Key Laboratory of Adolescent Cyberpsychology and Behavior (CCNU), Ministry of Education; School of Psychology, Central China Normal University; Key Laboratory of Human Development and Mental Health of Hubei Province, Wuhan 430079, China)
  • Received:2014-12-29 Published:2015-10-25 Online:2015-10-25
  • Contact: JIANG Guangrong, E-mail:



Working alliance in counseling has been a highly attended research area for several decades. Although a positive relationship between working alliance and counseling outcome has been well established, little do we know how working alliance functions in counseling process. Studying the development of working alliance would contribute to this understanding. Some theories have focused on describing developmental patterns of working alliance based on existing research findings, such as the U-shaped model and the rupture-repair episodes model. However, empirical effort has failed to show significant effect of these hypothesized developmental patterns on counseling outcomes. This naturalistic study aimed at exploring the developmental patterns of working alliance as emerged in counseling sessions and examining the relationship between these patterns and therapeutic outcomes.
The participants included 30 clients from 4 university counseling centers, 10 males and 20 females, with the number of sessions ranging from 4 to11, M = 6.53, SD = 2.03, Mdn = 6. By the time of data collection, all of the clients had finished their therapies. There were 20 therapists in the study, 3 males and 17 females, with professional experience from 1 to 22 years. They rated their therapeutic orientations using a 5-point Likert scale, which resulted in the strength of endorsement, from high to low, on person-centered, cognitive-behavioral and psychodynamic approaches. Each therapist offered 1 to 3 cases respectively. Clients filled out measurements of working alliance and depression symptoms after each session. Four shape-of-change parameters (Stiles et al., 2004) were used to describe the developmental patterns of working alliance in the counseling process of these 30 cases.

Results from the cluster analysis revealed three developmental patterns, labeled as linear increase, linear decrease, and quadratic increase. However, none of these developmental patterns had significant effects on the counseling outcomes. Then, correlations between the four shape-of-change parameters (i.e., indicators of developmental patterns) and outcomes were tested, but none of the correlations was significant either. It was also found that the rupture-repair episodes defined in terms of various criteria could not differentiate good outcomes from poor ones. Notably, the levels of working alliance were still found to predict outcomes. In order to explore the reasons why the developmental patterns had no relationship with outcomes, we conducted case studies comparing the working alliance developmental patterns in cases with good or poor outcomes. Results showed that the same developmental patterns emerged in both types of cases, but these pattern may have different meanings for different clients. It appeared that therapists’ regulations of working alliance in early sessions may have great influence on therapeutic outcomes.

Key words: working alliance, developmental patterns, outcomes, counseling