A longitudinal investigation of the relationship between Chinese counseling trainees’ anxiety and client symptom outcome

ABSTRACT

Objective

This study examined the reciprocal association between counseling trainees’ trait and state anxiety and their clients’ symptom distress and the mediating effects of working alliance.

Method

Data set included 6,888 sessions conducted by 211 master’s level beginning therapists with their 1,216 clients in China. Trainees completed a measure of trait anxiety at the beginning of practicum. Before each session, trainees completed the state anxiety measure and clients reported their symptom distress; and after every session, both the trainees and clients rated their perceived working alliance.

Results

Multilevel modeling showed that (a) Therapist trainees’ level of trait anxiety did not predict overall client outcome in terms of symptom reduction. (b) At the session-to-session level, higher therapist state anxiety before one session significantly predicted higher client distress before the next session, and higher client distress before one session significantly predicted higher therapist state anxiety before the subsequent session. (c) Therapist and client perceptions of working alliance were both significant mediators of the reciprocal relationship between therapist state anxiety and client distress.

Conclusions

Findings suggested that it was trainees’ state anxiety, instead of trait anxiety, that reciprocally related to client symptom outcome at the session-to-session level. Implications for clinical training were discussed.

PRACTICAL IMPLICATIONS

• A generally more anxious therapist trainee does not seem to have worse overall clinical effectiveness compared with a generally less anxious trainee.

• From session to session, trainees who experience less state anxiety tend to establish a stronger working alliance with their clients, which leads to more client symptom relief.

• Evidence suggests against selecting trainees based on their trait anxiety level; but it is recommended that supervisors help trainees manage their state anxiety before a session.

KEYWORDS:

• Therapist anxiety
• client distress
• working alliance
• longitudinal model
• dynamic systems

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09515070.2023.2212593

Notes

1. We set the minimum number of Bayesian iterations to 10,000 with a thinning parameter of 10 (essentially running 100,000 iteration cycles). For model convergence and fit, we used the criterion of Potential Scale Reduction (PSR) value being smaller than 1.05 for the last half of all iterations to ensure proper model convergence and fit (Asparouhov & Muthén, 2010.

2. Here we used the proportion of indirect effect in the overall effect (all indirect paths plus the direct path) to represent the effect size (ES) of the indirect mediation path. It ranges from 0 to 1 (100%) and shows the percentage of total effect accounted for by this particular mediation path. Same for below where indirect effect sizes are calculated.