This study employs latent growth mixture modeling (LGMM) to investigate the heterogeneous developmental trajectories of listening competency among college students participating in a peer psychological counseling course, and examines the predictive roles of Big Five personality traits and coping styles. Peer psychological counseling constitutes a form of non-professional helping, where training of listening competency should precede the development of psychological intervention competency. Listening competency refers to the process of receiving and constructing meaning from verbal and non-verbal messages, and then responding appropriately. While existing research demonstrates that integrated training combining theoretical instruction, case discussions, and experiential exercises can enhance listening competency, most studies operate under a homogeneity assumption, presuming uniform change trajectories across all trainees. Drawing on a person-centered perspective and empirical evidence from counselor training, which reveals diverse developmental pathways, this study proposes that students exhibit multiple distinct latent trajectory classes (H1). Furthermore, based on the broad influence of personality, the transactional model of stress and coping and relevant evidences from counselors training researches, it is hypothesized that Big Five personality traits (H2) and coping styles (H3) predict trajectory class membership.

A longitudinal design was employed with assessments at three time points: pre-course (T1), immediately post-course (T2), and 3-month follow-up (T3). A total of 206 undergraduates completed a battery of measures, including the Psychological Monitors' Listening Competency Questionnaire, the Chinese Big Five Personality Inventory and the Simplified Coping Style Questionnaire. The 8-week course integrated online theoretical modules with offline sessions featuring case discussions and experiential exercises. Data were analyzed using repeated-measures ANOVA to assess overall change, LGMM to identify unobserved trajectory classes, and multinomial logistic regression to examine predictors of subgroup membership.

The results showed that: (1) The main effect of time on listening competency was significant (F(1.55,318.36) = 379.76, p < 0.001, ηp²= 0.65, BF10 = 1.65×10^25). Post hoc comparisons using the Bonferroni method indicated that listening competency scores at T2 were significantly higher than those at T1 (MD = 1.11, SE = 0.05, p < 0.001, BF10,U = 2.97×10^21). Scores at T3 were significantly lower than those at T2 (MD = -0.85, SE = 0.05, p < 0.001, BF10,U = 0.16), but remained significantly higher than those at T1 (MD = 0.27, SE = 0.03, p < 0.001, BF10,U = 1.35×10^14). (2) LGMM identified four distinct trajectory classes: a Minimal-Change group (29.13%), a Continuous-Improvement group (11.65%), a Benefit-Maintenance group (37.38%), and a Post-Course Retreat group (21.84%), supporting H1. Post-hoc comparisons of listening competency scores at each time point confirmed that the four classes differed significantly and that their change patterns were consistent with the assigned labels. (3) Trajectory classification was significantly predicted by dimension of Big Five personality traits and coping styles, supporting H2 and H3. Precisely, the Benefit-Maintenance group was predicted by higher conscientiousness (χ²(1) = 5.88, p = 0.015, OR = 1.91) and higher positive coping (χ²(1) = 11.43, p = 0.001, OR = 8.43). The Post-Course Retreat group was predicted by lower openness (χ²(1) = 3.86, p = 0.05, OR = 0.64) and higher positive coping (χ²(1) = 9.14, p = 0.003, OR = 8.18).

This study concludes that the developmental trend of listening competency in response to peer counseling training is not uniform but exhibits significant heterogeneity, encompassing multiple distinct change trajectories. The findings challenge the assumption of a single average course training effect and underscore the critical influence of individual differences and instructional perceptions on both immediate learning and long-term retention. The application of LGMM provides a novel methodological contribution, enabling the identification of at-risk subgroups, such as those prone to skill decay. However, it should be noted that the overall predictive power of Big Five personality traits and coping styles was relatively low. Future research should explore additional variables highlighted in the counselor training literature, such as trainees' self-doubt, attachment styles, cultural humility, and supervision style matching. Besides, potential confounds like instructional consistency and student attendance should also be controlled for.

Practically, these results suggest that trajectory of changes in college students' listening competencies during peer counseling courses varies from person to person. Therefore, instructors should provide early personalized training for those showing "minimal-change", offer consolidation exercises for those exhibiting signs of "post-course retreat", and give affirmation and praise to those demonstrating "continuous-improvement" or "benefit-maintenance". At the same time, these results further suggest that personalized training approaches, early identification of students with specific trait profiles (e.g., low conscientiousness, low openness), and efforts to enhance the perceived value of all course components are essential for improving the efficacy and sustainability of peer counselor education programs.