Content area
Network analysis is widely used in psychological research, yet the applicability in the counseling process of novice psychologists remains unclear. This study applies network analysis to explore lexical changes in novice psychological counselors over four consecutive sessions in order to understand the growth process of novice counselors. We analyzed verbatim transcripts of 221 novice counselors using a psychological dictionary to convert them into word frequencies. Descriptive statistical analysis was conducted on personal pronouns, emotional process words, cognitive process words, and personal concern words. Network analysis was then applied to assess the structures of these word frequencies. Finally, correlation analysis was conducted between novice counselors’ and clients’ word frequencies. Across four sessions, self-related personal pronouns in novice counselors increased, while positive emotion and anxiety words decreased. Causal, discrepancy, and pause words decreased in cognitive processes, while physiological process words increased in personal concerns. Network analysis showed increasing block-style connections among words used by counselors. Correlation analysis indicated decreasing consistency in word frequencies between counselors and clients. Network analysis is effective for studying novice counselors’ processes, providing insights into lexical changes and relationships between counselors and clients.
Introduction
Novice counselors who were transited from amateur helpers to professional practitioners through formal psychological training1 undergo a developmental process from novice to proficient and expert stages2. With the growing demand for psychological counseling in China, there is a significant need for qualified professionals. Enhancing both the quantity and quality of psychological counseling services is essential. Effective training methods for novice counselors are crucial to meet these demands and improve counseling outcomes3.
The psychological counseling process is dynamically oriented toward resolving clients’ psychological issues through verbal communication. Novice counselors must transition from the orientation phase to the procedural phase in their vocational tasks. While training can impart knowledge and skills, real-world problem-solving ability is honed through actual counseling sessions and diverse client interactions. This growth process is distinctive, personalized, and continuously evolving. Assessing counselors’ dynamic progression from novice to proficient is advantageous for designing targeted vocational training. It strengthens their core professional competencies and is essential for their personal development.How can we dynamically evaluate counselors` mastery of skills throughout the counseling process? Providing targeted supervision and feedback is essential for facilitating the growth of novice counselors. Researches have confirmed that the counselor’s impact on the effectiveness of counseling is often more significant than the counseling method itself. The counselor’s personality plays a key role in influencing clients and promoting change4,5.
Understanding the changes in word usage by novice counselors during their sessions is meaningful for their personal growth. Current research focuses on their development, performance of counseling skills, and various factors related to supervision and training.Many novice counselors experience disorientation during their initial learning phase, rendering continuous reflection and learning crucial for their development. Unlike proficient counselors, who demonstrate strong cognitive, emotional, and relational abilities, novices often face challenges in emotional acceptance, processing, and relationship-building6. They may excessively focus on negative behaviors, superficial cognition, anxiety, and feelings of incompetence, common challenges faced by many novice counselors. Tériault, Gazzola, & Richardson (2009) believe that feelings of incompetence are a significant characteristic of the professional development process for novice counselors7. Additionally, Jennings & Skovholt (1999) found that novice counselors prioritize establishing the counseling relationship during their initial sessions while facing limitations in cognitive, emotional, and relational performance8.
Verbal organization
Novice counselors often use isolated and fragmented verbal responses9, rather than cohesive or spiral verbal techniques4. This tendency can hinder their ability to explore counseling deeply, maintain structural control, and effectively guide clients in establishing meaningful tasks and goals. In contrast, proficient counselors are more adept at using chunked speech. A study by Xiao (2016) indicates that proficient counselors typically employ chunked and spiral speech response techniques, allowing them to navigate the counseling structure effectively, establish emotional connections with clients, and guide them toward consistent tasks and goals9. Proficient counselors are more inclined than novices to use chunked speech when asking questions about insight and emotional feelings4. Based on the literature and empirical studies mentioned above, we propose the following hypothesis:
H1 During the growth process, novice counselors will increasingly employ block-like and spiral verbal responses.
Psychological counseling style
Novice counselors may lean toward didactic approaches, relying on their personal values and judgments, which can lead to overlooking clients’ emotional states2. Research by An et al. (2015) indicates that novice counselors often use their own values and criteria for judgment, such as what is considered “normal” or in need of improvement1. This approach can make counselors appear "preachy," causing clients to feel misunderstood or rejected. In their eagerness for clients to accept their views, novice counselors may inadvertently overlook the clients’ emotions. Thus, we hypothesize that:
H2 During the growth process, the didactic form of verbal content used by novice counselors will gradually decrease.
Empathy
Empathy is a critical predictor of therapeutic success10, and often poses a challenge for novice counselors who struggle to establish emotional resonance with their clients. Research indicates that higher levels of empathy enhance clients’ perceptions of being understood, leading to improved treatment outcomes. However, achieving high levels of empathy can be particularly challenging for novice counselors2. Early in their careers, these counselors often experience weak emotional alignment with clients, making it difficult to form meaningful emotional connections9. Reflecting on developing empathetic conversational patterns was crucial for these counselors11. Greenberg et al. (2012) discovered that greater empathy was linked to clients feeling more understood and enjoying better therapeutic outcomes12.Counselors who demonstrate high empathy typically have higher success rates in therapy, reinforcing empathy’s role as a crucial predictor of therapeutic effectiveness10. Ni(2020)indicated that counselors’ emotional engagement is influenced by their experience, with more seasoned counselors exhibiting higher empathy levels4. In the early stages of counseling, proficient counselors show greater emotional congruence with clients compared to novices4. Proficient counselors showed greater emotional congruence with clients compared to novices13. Ronnestad and Skovholt emphasized the importance of understanding the empathetic conversation patterns of novice counselors, as this understanding could facilitate their reflective learning and development. Therefore, we propose the following hypothesis:
H3 During the growth process, novice counselors will show more empathy and pay increasing attention to the emotional responses of clients.
Linguistic synchrony with clients
Current research on linguistic synchrony has primarily focused on non-verbal aspects, such as actions. Studies suggested that moderate levels of non-verbal synchrony were more effective in improving client symptoms14. However, there is limited research on the linguistic consistency characteristics of counseling sessions conducted by novice counselors15,16.
Paulick et al. (2018) categorized clients based on therapeutic outcomes into groups exhibiting varying levels of motor synchrony14. The improvement group exhibited moderate motor synchrony, while the no improvement–dropout group showed low levels, and the no improvement–mutual voluntary termination group demonstrated high levels14. An ideal level of non-verbal synchrony should be moderate; excessive synchrony may be perceived as imitation, while insufficient synchrony can appear insincere16. Maintaining moderate non-linguistic synchrony and consistency with clients can enhance counseling effectiveness.
Given the deficiencies observed in novice counselors, mastering the lexical characteristics of their counseling is essential. Targeted professional guidance and training are necessary to address these deficiencies. Hill et al. (2007) argued that clarifications, suggestions, and feedback from supervisors were vital for the growth of novice counselors13. Furthermore, Hill et al. (2008) found that trainees who received training employed more exploratory techniques, demonstrated greater empathy, spoke less, and felt a higher sense of self-efficacy17. Wang, Jia & Zhang (2015) discovered that supervision significantly enhances the professional learning and development of novice counselors, which in turn fosters linguistic synchrony with clients18. Thus, we hypothesize that:
H4 During the growth process, the verbal congruence between counselors and clients should begin to slowly decrease as the counseling sessions deepen.
Current study
Network analysis was used to visualize the lexical characteristics and growth in novice counselors’ counseling sessions. This structural model, composed of nodes and edges19, has been applied in various psychological fields20,21. It provides an intuitive reflection of the characteristics and challenges faced by novice counselors and elucidates the dynamic relational changes between counselors and clients.
Materials and methods
Participants
The study utilized purposive sampling to select novice psychological counselors based on specific inclusion criteria: (1) completion of basic psychology coursework and training in core psychological counseling skills (psychological assessment, diagnosis, and counseling); (2) conducting their first individual counseling sessions, consisting of at least three sessions each counselor; (3) adherence to professional ethical standards, with sessions recorded with client consent; (4) confirmed counseling effectiveness rated as satisfactory or higher by clients; (5) counselors utilizing fundamental skills such as counseling reception techniques, relationship-building techniques (empathy, active listening, respect, warmth, sincerity), and common conversational techniques (listening, questioning, encouragement, repetition, specification, immediacy, facial expression, content response, emotional response, observation, summary, explanation, guidance, content expression, emotional expression, self-disclosure, influential summary, non-verbal communication, participatory and influential techniques), without exclusive reliance on any specific technique; (6) clients seeking counseling primarily for developmental or health consultation issues within normal ranges, excluding consultations for disorders beyond this range22.
Novice psychological counselors must complete at least three years of professional training, demonstrating competence in basic knowledge and skills. Consultations with clients should last approximately 50 min (ranging from 35 to 65 min), once or twice a week, face-to-face in a standard counseling room. The novice psychological counselors in this study, aged 24 to 38 (M = 30.23,SD = 0.36 ), were all Han Chinese nationals and came from universities or psychological training institutions with counselor-training qualifications. They systematically studied counseling methods and techniques, learned to identify and assess psychological problems, and participated in a three-year professional training program. Ultimately, 221 participants were enrolled, including 58 males and 163 females.
The study was approved by the Ethics Committee of Changsha Social Work College (No. 202411) in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants after a detailed description of the procedures was provided.
Sample size estimation
Network analysis requires a sample size calculation23,24. The sample size must exceed the total parameters, including threshold parameters and pairwise association parameters. For threshold parameters, this equals the number of nodes, and for pairwise association parameters, it equals the number of nodes multiplied by (number of nodes—1)/225. In this study, we selected 18 word frequency categories that served as the nodes, including personal pronouns (Self—reference, Family, Friend), emotional process words (Positive Emotion, Negative Emotion, Anxiety, Anger, Sadness), cognitive process words (Insight, Causation, Discrepancy, Tentative), and personal concern words (Bio, Work, Achievement, Leisure, Money, Death). These 18 word frequency categories were chosen as research nodes since the validity analyses showed that they can accurately reflect psychological traits26. This study established 18 nodes, leading to 18 threshold parameters and 153 pairwise association parameters (using the formula 18 × 17/2). Consequently, a minimum sample size of 171 participants was determined to be necessary. The actual inclusion of 221 participants was deemed sufficient for the study’s requirements.
Instruments
Text analysis tools
The language Inquiry and Word Count (LIWC) System was used for text analysis, allowing for automated batch processing and incorporating a psychological dictionary27. Internal consistency of LIWC categories ranges from 0.02 to 0.7528. The Simplified Chinese version (SC-LIWC) shows similar consistency to foreign versions29. LIWC has been widely used in psychotherapy and psychological counseling research30, 31, 32, 33, 34–35.
Text Mind 4.0 (Wenxin), developed by the Computational Psychology Laboratory of the Institute of Psychology in the Chinese Academy of Sciences, is a Chinese psychological analysis system that extracts characteristic words from transcripts36. It has been validated for frequency analysis of Simplified Chinese characters26,37.
Text classification tool
A Java-based self-programmed tool was used for text classification, segmenting transcripts into independent texts.
Procedure
In this study, novice counselors conducted continuous consultations with assigned clients, without substitutions during the process. They held one or two sessions per week in a standardized counseling room to ensure privacy. Clients consented to session recordings, which were transcribed verbatim by counselors. Counselors annotated transcripts with details on techniques, methods, and their counseling reflections. Transcripts with over 20,000 words or under 5,000 words per session were excluded.
After each session, supervisors reviewed the transcripts and provided guidance in face-to-face supervision sessions. Each novice counselor worked with a personal supervisor with over 15 years of experience. Counselors submitted verbatim transcripts of at least three consecutive sessions and received supervision at least three times in the last six months, with the same supervisor consistently providing oversight.
Data analysis
Descriptive statistical analysis of word frequencies was conducted using SPSS 25.0, with results presented as mean ± standard deviation. Normality tests confirmed a normal distribution for word frequencies. Pearson correlation analysis was assessed between counselors and clients across four counseling sessions.
Network analysis which is a visualization-based method for exploring relationships and interactions among different psychological factors through nodes and lines was used in the study. Network structures can be built via correlation matrix or partial correlation networks. Centrality indices quantify a node’s significance within a network structure. Three common types are: (1) Strength: the number of nodes directly connected to a given node. (2) Closeness: the sum of the shortest distances from a node to all others. (3) Betweenness: the frequency with which a node lies on the shortest paths between other pairs of nodes. A node with greater strength is more centrally positioned and more important to the network. Higher closeness indicates that the node is closer to others and more centrally located. Betweenness reflects a node’s mediating power over interactions between other nodes. For psychological networks, strength is considered the most relevant measure. Network stability reflects its ability to retain fundamental structure and function amid internal or external disturbances. We also employed the correlation stability coefficient (CS coefficient) which assesses the correlation between estimation results from the full sample and progressively smaller samples. Typically, a CS coefficient of 0.7 is deemed the minimum acceptable threshold, while values below 0.25 are considered unsatisfactory. For robust analysis, coefficients should ideally exceed 0.5. The network comparison test (NCT) was used to compare networks based on their structure and properties38, aiming to identify similarities and differences39. In conventional network comparisons, visually assessing differences between network structures is highly subjective and lacks objective quantification. To address these limitations, researchers have proposed using permutation testing as a statistical method for directly comparing different network structures40.
Network analysis was performed using the qgraph package in R 4.2.0, calculating network structure components and centrality indices41. The bootnet package was used to assess the stability of key centrality indicators, and the Network Comparison Test package compared different network parameters.
Results
Word frequency descriptive statistics of counselors across 4 sessions
Different frequencies reflect distinct psychological meanings: Personal pronouns indicate psychological focus and social relationships, emotional words convey personal emotions, and cognitive process words illustrate thinking styles42,43. These frequencies have been shown to be effective and closely associated with psychological traits26,44,45. To account for the decimal nature of these frequencies, each original frequency value was multiplied by 100 after extraction.
During the four counseling sessions (see Table 1), the use of self-related words by counselors and clients indicated that counselors increasingly focused on self-reference content, while clients consistently concentrated on their own experiences. The use of family and friend words declined over time, suggesting reduced attention to these topics as counseling progressed. In terms of emotional words, both counselors and clients utilized more positive emotion words in the fourth session, while the frequency of negative emotion and sadness words remained low and stable. Counselors used more anxiety words in the third session but decreased their use thereafter, while clients’ usage remained consistent. Both parties infrequently used angry words, with little variation noted. Additionally, counselors’ use of insight decreased, whereas clients’ use increased. The frequency of personal concern words varied between counselors and clients. Overall, these changes reflect dynamic shifts in focus throughout the counseling process.
Table 1. Frequency descriptive statistics of counselor’s 4 counseling sessions (M ± SD).
Variables | Time1 | Time2 | Time3 | Time4 | ||||
|---|---|---|---|---|---|---|---|---|
CO | CL | CO | CL | CO | CL | CO | CL | |
Self-reference | 1.37 ± .75 | 4.15 ± 1.24 | 1.53 ± .94 | 4.16 ± 1.39 | 2.19 ± .06 | 3.94 ± 1.41 | 3.89 ± 2.49 | 3.90 ± 1.44 |
Family | .61 ± .73 | .59 ± .60 | 1.80 ± .02 | .58 ± .63 | .54 ± .66 | .37 ± .52 | .50 ± .68 | .35 ± .47 |
Friend | .29 ± .26 | .23 ± .19 | .22 ± .25 | .20 ± .23 | .06 ± .09 | .19 ± .20 | .09 ± .08 | .18 ± .22 |
PosEmo | 1.74 ± .63 | 1.64 ± .53 | 1.67 ± .57 | 1.76 ± 1.02 | 1.57 ± .16 | 2.05 ± 1.61 | 1.41 ± 0.67 | 2.19 ± 1.37 |
NegEmo | 1.12 ± .51 | 1.22 ± .45 | .52 ± .12 | 1.19 ± .50 | 1.23 ± .15 | 1.19 ± .64 | .96 ± .38 | 1.14 ± .62 |
Anx | .27 ± .27 | .23 ± .19 | .12 ± .05 | .22 ± .20 | .66 ± .34 | .27 ± .46 | .06 ± .02 | .27 ± .31 |
Anger | .28 ± .19 | .32 ± .19 | .11 ± .09 | .31 ± .21 | .18 ± .22 | .29 ± .27 | .31 ± .01 | .23 ± .21 |
Sad | .18 ± .16 | .19 ± .13 | 0 | .19 ± .16 | .25 ± .03 | .21 ± .19 | .14 ± .01 | .29 ± .27 |
Insight | 2.76 ± .74 | 2.33 ± .65 | 3.08 ± .81 | 2.33 ± .89 | 1.73 ± .70 | 2.44 ± .83 | 2.21 ± 0.87 | 2.89 ± 1.01 |
Cause | 1.39 ± .49 | 1.19 ± .35 | 1.46 ± .15 | 1.18 ± .47 | 1.11 ± 0.17 | 1.15 ± .49 | 1.06 ± 0.28 | 1.19 ± .58 |
Discrep | 2.72 ± .73 | 2.53 ± .69 | 3.22 ± .28 | 2.69 ± .86 | 3.42 ± 0.26 | 2.85 ± .97 | 2.27 ± 0.13 | 2.68 ± 1.03 |
Tentat | 3.42 ± .83 | 2.99 ± .76 | 3.76 ± .42 | 3.14 ± .93 | 3.25 ± .01 | 3.33 ± .97 | 2.82 ± 0.11 | 3.35 ± 1.07 |
Bio | .89 ± .54 | 1.32 ± .61 | .61 ± .26 | 1.36 ± .65 | 1.41 ± .22 | 1.41 ± .78 | 2.25 ± 0.37 | 1.41 ± .87 |
Work | 2.47 ± 1.12 | 2.23 ± 1.06 | 1.70 ± 0.66 | 1.96 ± 1.07 | 2.45 ± 1.12 | 1.77 ± 1.05 | 2.11 ± 1.32 | 1.69 ± 1.20 |
Achieve | 1.14 ± .52 | .75 ± .38 | 1.10 ± 0.21 | .72 ± .44 | 1.19 ± .27 | .72 ± .46 | .37 ± .06 | .80 ± .57 |
Leisure | .49 ± .31 | .61 ± .30 | .36 ± .17 | .62 ± .39 | .54 ± .45 | .65 ± .41 | .46 ± .00 | .55 ± .50 |
Money | .13 ± .15 | .17 ± .22 | .08 ± .11 | .21 ± .28 | .12 ± .17 | .21 ± .31 | .18 ± .05 | .19 ± .31 |
Death | .04 ± .08 | .06 ± .08 | .06 ± .08 | .07 ± .09 | .01 ± .02 | .06 ± .09 | 0 | .063 ± .17 |
CO counselors, CL clients.
Network structures of counselors and clients across 4 counseling sessions
The network comparison test (NCT) is significant in research as it sharply reflects changes in the consultation content and approach of novice counselors across four counseling sessions. Comparing the four networks between novice counselors and clients reveals gradual changes in their interaction dynamics as the consultation content deepens. Figures 1, 2 illustrate the network structures of counselors and clients, showing a modular increase in connectivity over the four sessions. This modularity is especially evident in the emotional process word clusters, which are more pronounced in client networks. Novice counselors show a spiral—like progression in network connections from the first to the fourth session. By the fourth session, their networks exhibit enhanced interconnections, mirroring the node relationships of the third session but with greater connectivity. Clients, however, display more marked changes, particularly in the fourth session, where interconnections among network nodes increase significantly.
Fig. 1 [Images not available. See PDF.]
Network diagram of 4 counseling sessions by counselors.
Fig. 2 [Images not available. See PDF.]
Network diagram of 4 counseling sessions by clients.
Correlation analysis of counselors’ and clients’ word frequencies across counseling sessions
In the four counseling sessions (see Table 2), family and friend words showed high consistency between counselors and clients, indicating that these topics are significant and often discussed in clients’ lives. For emotional words, the correlation of positive emotions decreased, while negative emotion correlation remained stable and strengthened by the fourth session. Anxiety correlation spiked in the fourth session, anger correlation fell, and sadness correlation rose. Among cognitive and insight-related words, the correlation of insight increased by the fourth session, while correlations for causation, discrepancy, and tentative words declined. Self-reference words had a low correlation, even turning negative in the fourth session. The correlation coefficients of other personal concern words fluctuated, reflecting shifting focal points, with death words consistently showing low correlation. Overall, these changes reveal dynamic shifts in focus during counseling, with adjustments in consistency across various topics over time and changing attention to specific emotions and cognitive concepts.
Table 2. Correlation coefficients of word usage between counselors and clients in 4 counseling sessions.
Variables | Time1 | Time2 | Time3 | Time4 |
|---|---|---|---|---|
Self-reference | 0.05 | 0.10 | 0.27** | −0.08 |
Family | 0.77** | 0.81** | 0.74** | 0.62** |
Friend | 0.61** | 0.61** | 0.41** | 0.34* |
PosEmo | 0.46** | 0.34** | 0.22** | 0.10 |
NegEmo | 0.54** | 0.52** | 0.43** | 0.55** |
Anx | 0.59** | 0.50** | 0.45** | 0.78** |
Anger | 0.45** | 0.37** | 0.33** | 0.08 |
Sad | 0.49** | 0.39** | 0.39** | 0.50** |
Insight | 0.47** | 0.32** | 0.32** | 0.44** |
Cause | 0.31** | 0.34** | 0.35** | 0.23 |
Discrep | 0.36** | 0.38** | 0.28** | 0.25 |
Tent | 0.38** | 0.25** | 0.13 | 0.21 |
Bio | 0.61** | 0.50** | 0.38** | 0.69** |
Work | 0.70** | 0.69** | 0.55** | 0.50** |
Achieve | 0.61** | 0.62** | 0.55** | 0.67** |
Leisure | 0.51** | 0.48** | 0.41** | 0.53** |
Money | 0.60** | 0.67** | 0.74** | 0.56** |
Death | 0.55** | 0.42** | 0.42** | 0.20 |
n | 221 | 221 | 221 | 37 |
*P < 0.05, **P < 0.01.
Results of network centrality in counseling sessions
Figures 3, 4 present centrality measures (strength, closeness, and betweenness) for the networks across sessions. By the fourth session, both counselors and clients exhibit peak strength and closeness centrality. This indicates more intense interactions (higher strength) and denser network structures (higher closeness). These metrics highlight key nodes in the network, reflecting the intensified and evolving nature of counselor-client interactions, particularly evident in the fourth session.
Fig. 3 [Images not available. See PDF.]
Centrality in 4 sessions of counselor consultation.
Fig. 4 [Images not available. See PDF.]
Centrality in 4 sessions of client consultation.
Stability of network results
Figure 5 presents that at a 95% confidence level, the stability coefficients of the expected effects are all above 0.75, suggesting high stability. This indicates that the centrality indices are robust and consistent across different data subsets, thus reinforcing the reliability and validity of the network analysis findings. The stability observed underscores the dependability of the centrality metrics in assessing counselor-client interactions throughout the counseling sessions.
Fig. 5 [Images not available. See PDF.]
Stability estimates of counselors’ 4 counseling sessions.
Discussion
This study integrates qualitative and quantitative approaches to investigate the language use of novice counselors and its evolution throughout counseling sessions. The results uncover several patterns in the word usage changes of novice counselors, which are elaborated below.
Regarding hypothesis H1, it suggests that while novice counselors start with a focus on cognitive insights, they gradually learn to incorporate a more holistic approach that includes emotional understanding, while novice counselors initially focus on cognitive insights, they gradually learn to incorporate a more holistic approach that includes emotional understanding, which is vital for effective counseling. This progression aids not only in resolving client issues but also contributes to the personal development of the counselors. Novice counselors increasingly use self-related personal pronouns, indicating a growing awareness of their clients’ self-states29. This self-focus is beneficial for the counselors’ personal growth. Initially, the language used by novice counselors is dominated by cognitive process words, indicating a focus on cognitive insights, often involving discussions of facts or pauses in conversation. By the fourth session, there is a noticeable increase in the use of first-person and personal concern words, suggesting that counselors are beginning to address clients’ personal concerns more directly, which is crucial for effective counseling.The high frequency of cognitive process and personal concern words may indicate that novice counselors focus too much on factual issues presented by clients, possibly overlooking underlying emotions and deeper meanings in their narratives. This could lead to clients feeling that their emotional states are not fully understood. The increase in emotional process words in the fourth session indicates that novice counselors are beginning to shift their attention toward the emotional and personal concerns of their clients, essential for establishing a supportive therapeutic relationship. The decrease in family and friend words compared to self-reference words may be due to several factors. First, the limited number of counseling sessions may restrict discussions, causing clients to concentrate more on their internal psychological experiences and emotional feelings. Thus, they may use more self-referential words for in-depth exploration, while discussions about family and friends decline. Second, novice counselors are still developing their counseling self-efficacy. When addressing multicultural issues presented by clients, they may lack the experience to explore the influence of family and friends on clients’ adaptation. Consequently, the frequency of family and friend words is lower than that of self-reference words, reflecting limitations in uncovering significant interpersonal relationships in clients’ multicultural contexts. Enhancing their counseling self-efficacy can improve this situation, allowing family and friend-related factors to be more fully reflected in counseling dialogues.
The second hypothesis in our study was also conformed. While novice counselors may begin with less structured responses, they show progress over time, and targeted training could further enhance their ability to connect emotionally with clients and provide more effective counseling9. Network analysis indicates that novice counselors initially have weak or nonexistent connections between the words they use, suggesting their verbal responses lack structure and cohesion. In contrast, experienced counselors demonstrate more structured and cohesive verbal responses, implying a higher level of skill and experience in counseling. Although the study only recruited novice counselors, changes across the four network analyses suggest that the network connections of novice counselors increased, with chain-like and block-like connections gradually emerging. This indicates a slow but noticeable transition towards proficiency as the number of counseling sessions increases. Over the course of four sessions, novice counselors begin to develop clustered connections between words, indicating an improvement in their counseling techniques. Nevertheless, compared to experienced counselors, their verbal response structures are still less systematic and comprehensive, indicating potential for further skill development.
The third hypothesis, posits that empathy, identified as a crucial skill for counselors, enables them to connect with clients on an emotional level and understand their needs46. The findings reveal that over time, novice counselors are altering their word usage, placing greater emphasis on clients’ emotional states and personal concerns. This shift signifies an enhancement in empathy among novice counselors, as they grow more adept at comprehending and addressing the emotional facets of their clients’ experiences. Such alignment of counselors’ responses with clients’ viewpoints is essential for fostering trust and rapport within the counseling relationship.
Hypothesis 4 posits that the consistency in word usage between counselors and clients is not static but evolves as the counseling process progresses. Initially, consistency helps establish rapport; however, as counselors delve deeper into clients’ issues, they may use language that challenges and clarifies, leading to a decrease in consistency. This reflects a mature and responsive approach to counseling, where the counselor’s language adapts to the client’s needs and therapeutic goals. At the beginning of the counseling process, counselors may use a consistent level of word usage to establish rapport with their clients. This consistency can help create a comfortable environment and build trust. As counseling progresses, counselors may begin to challenge and clarify clients’ values and proposed solutions to their problems. This can involve questioning and negation, which is a natural part of the therapeutic process as counselors help clients explore different perspectives and consider alternative solutions. The study suggests that the consistency in word usage between counselors and clients tends to decrease as the number of counseling sessions increases. This decrease in consistency is not necessarily negative; rather, it reflects the evolving nature of the counselor-client relationship. The decrease in word usage consistency is attributed to counselors engaging more deeply with clients’ issues and solutions over time. As counselors become more familiar with the clients’ perspectives and needs, they may adjust their language to better address those specific concerns. The evolving nature of counselor-client interactions is highlighted, indicating that the relationship is dynamic and changes as both parties work together to address the clients’ concerns. This process of decreasing consistency can also be seen as part of the professional development of the counselor, as they learn to adapt their communication style to better suit the needs of individual clients.
The network analysis shows high stability coefficients, which strongly confirm the reliability of the methodology. This validation ensures the network analysis results are reliable and the centrality metrics for counselor-client interactions are trustworthy. Stability coefficients in network analysis measure the consistency of results across different samples or iterations. High coefficients demonstrate the robustness of the network analysis methodology, ensuring consistent and accurate results. In this case, the high coefficients suggest the network analysis effectively captures counselor-client interaction dynamics. Centrality metrics in network analysis help identify important nodes like words or concepts. The high stability coefficients ensure these metrics reliably assess the centrality of different elements in the counselor-client interaction network.
This study’s results show that boosting counseling skills is significant in multiple aspects. For novice counselors, shifting focus from cognitive to emotional aspects enhances counseling effectiveness and care’s humanistic side. Better language structure and coherence can make counseling smoother, which is key for building stable therapeutic relationships and aiding client transformation. Heightened self—awareness can foster self—reflection and professional growth, while developed empathy can strengthen trust—based relationships and propel the counseling process. To close the gap with experienced professionals, novice counselors need to engage in continuous learning and practical application, consciously emulating exemplary cases while integrating their personal style and adapting their approach to clients’ needs at different counseling stages. Long—term tracking studies should be conducted to observe changes in novice counselors’ consultation skills, personal growth, and career development over time, identifying critical moments and influencing factors in their journey. Experimental research ought to be carried out to compare the effectiveness of various training methods in improving novice counselors’ consultation skills, offering empirical evidence for refining counselor training programs. It’s also essential to explore the correlation between counselors’ personal traits like empathy and self—efficacy and consultation outcomes, shedding light on how these traits impact the counseling process and results.
Limitations and future research directions
This study has several limitations. First, it only examined novice counselors, so no comparative data from experienced counselors are available. Second, the limited number of sessions and their set time frame restricted the depth of discussions. Third, variations in counselors’ professional abilities and cultural backgrounds influenced the results. Lastly, clients’ cultural contexts and other factors affected their word usage patterns. Future research should expand the sample to include experienced counselors and diverse participants. The study context can be optimized by increasing the number of sessions, extending their duration, and implementing standardized procedures. To gain a comprehensive understanding of the counseling process, it will be essential to conduct an in-depth analysis of individual differences between counselors and clients. Designing comprehensive training programs is also essential for improving novice counselors’ skills.
Conclusion
Network analysis of novice counselors’ counseling processes reveals dynamic growth in language use and counselor-client interactions. By analyzing four counseling sessions and examining correlations between counselors and clients, the study sheds light on the evolution of word usage consistency and the progression of interactions. These insights highlight how clients’ growth and problem-solving are reflected in their counseling sessions.
Author contributions
Zhiyong Wu designed the study, collected and analyzed data, and drafted the manuscript. Haiyan Ding collected data and conducted preliminary analysis. Xiaotong Tang conceived and designed the study. All authors approved the final manuscript.
Funding
Supported by the Hunan Provincial Social Science Achievement Evaluation Committee Project (XSP2023JYC108).
Data availability
The datasets generated and/or analyzed during this study are not available but may be requested from the corresponding author.
Declarations
Competing interests
The authors declare no competing interests.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1. An, Q; Jia, X; Dai, Y et al. Professional skills performance of novice counselors in their first counseling practice. Ch. J. Mental Health; 2015; 29,
2. Xu, H; Xie, L. Research on vocational and technical education in Germany. Beijing Normal Univ. Press; 2021; 11, 130.1:CAS:528:DC%2BB3MXnslyrsrs%3D
3. Skovholt, TM; Ronnestad, MH. The journey of the counselor and therapist: research findings and perspectives on professional development. J. Career Dev.; 2003; 30,
4. Ni, C et al. Interpersonal complementarity in counseling sessions and its relationship with working alliance and counseling effectiveness. J. Psychol.; 2020; 52,
5. Ni, J; Hou, Z. The trajectory of behavioral reflection in novice counselors. Chin. J. Clin. Psychol.; 2014; 2, pp. 367-372.
6. Jennings, L; Goh, M; Skovholt, TM; Hanson, M; Banerjee-Stevens, D. Multiple factors in the development of the expert counselor and therapist. J. Career Dev.; 2003; 30, pp. 59-72. [DOI: https://dx.doi.org/10.1177/089484530303000104]
7. Tériault, A; Gazzola, N; Richardson, B. Feelings of incompetence in novice therapists: Consequences, coping, and correctives. Can. J. Couns.; 2009; 43,
8. Jennings, L; Skovholt, TM. The cognitive, emotional, and Relational characteristics of master therapists. J. Couns. Psychol.; 1999; 46,
9. Xiao, T. The Relationship Between counselors’ verbal Response Techniques and Working Alliance in First-Time Psychological Counseling; 2016; Yunnan Normal University:
10. Moyers, TB; Miller, WR. Is low therapist empathy toxic?. Psychol. Addict. Behav.; 2012; 27,
11. Skovholt, TM; Ronnestad, MH. Struggles of the novice counselor and therapist. J. Career Dev.; 2003; 30,
12. Greenberg, LS. Emotions,the great captains of our lives: Their role in the process of change in psychotherapy. Am. Psychol.; 2012; 67,
13. Hill, CE; Sullivan, C; Knox, S; Schlosser, LZ. Becoming psychotherapists: Experiences of novice trainees in a beginning graduate class. Psychother. Theory Res. Pract. Training; 2007; 44,
14. Paulick, J; Deisenhofer, A-K; Ramseyer, F; Tschacher, W; Boyle, K; Rubel, J; Lutz, W. Nonverbal synchrony: A new approach to better understand psychotherapeutic processes and drop-out. J. Psychother. Integr.; 2018; 28, pp. 367-384. [DOI: https://dx.doi.org/10.1037/int0000099]
15. Moulder, RG; Boker, SM; Ramseyer, F; Tschacher, W. Determining synchrony between behavioral time series:An application of surrogate data generation for establishing falsifiable null-hypotheses. Psychol. Methods; 2018; 23, pp. 757-773. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29595296][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6163103][DOI: https://dx.doi.org/10.1037/met0000172]
16. Dai, X; Hu, Y; Zhang, Y. The phenomenon of interpersonal synchronicity: a new perspective for exploring alliance in the counseling process. Adv. Psychol. Sci.; 2022; 30,
17. Hill, CE; Roffman, M; Stahl, J; Friedman, S; Hummel, A; Wallace, C. Helping skills training for undergraduates: Outcomes and prediction of outcomes. J. Couns. Psychol.; 2008; 55,
18. Wang, W; Jia, X; Zhang, M. The supervision experience of novice counselors. Chin. J. Mental Health; 2015; 29,
19. Cai, Y; Dong, S; Yuan, S; Hu, C. A network analysis model between variables and its application. Adv. Psychol. Sci.; 2020; 28,
20. Christensen, AP; Golino, HF; Silvia, P. A psychometric network perspective on the validity and validation of personality trait questionnaires. Eur. J. Personal.; 2019; 34, pp. 1095-1108. [DOI: https://dx.doi.org/10.1002/per.2265]
21. Bringmann, LF et al. What do centrality measures measure in psychological networks?. J. Abnorm. Psychol.; 2019; 128, pp. 892-903. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31318245][DOI: https://dx.doi.org/10.1037/abn0000446]
22. China Employment Training Technical Guidance Center & China Mental Health Association. National vocational qualification training tutorial: psychological counselor; 2015; Minzu Publishing Press:
23. Epskamp, S et al. Qgraph: Network visualizations of relationships in psychometric data. Stat. Softw.; 2012; 48,
24. Shang, B et al. A network analysis of the relationship between narrative dysphoria and cognitive emotion regulation strategies in community-dwelling older adults with chronic disease co-morbidities. Ch. J. Mental Health; 2024; 04, pp. 318-324.
25. Li, S; Zhang, S. Relationship between peer neglect and cyberbullying perpetration. Ch. J. Mental Health; 2024; 2, pp. 109-113.
26. Zhao, N; Jiao, D; Bai, S et al. Evaluating the validity of simplified chinese version of LIWC in detecting psychological expressions in short texts on social network services. PLoS ONE; 2016; 11,
27. Tausczik, YR; Pennebaker, JW. The psychological meaning of words:LIWC and computerized text analysis methods. Soc. Psychol.; 2010; 29,
28. Pennebaker, Francis, Booth. Linguistic Inquiry and Word Count. Lawrence Erlbaum Associates Mahwah Nj. (2012).
29. Zhang, X. LIWC: A text analysis tool based on discourse metrics. J. Southwest (Univ. Natl. Human. Soc. Sci. Ed.); 2015; 36,
30. Zimmermann, J; Brockmeyer, T; Hunn, M; Schauenburg, H; Wolf, M. First-person pronoun use in spoken language as a predictor of future depressive symptoms: Preliminary evidence from a clinical sample of depressed patients. Clin. Psychol. Psychother.; 2017; 24,
31. Ramírez-Esparza, N., Chung, C. K., Kacewicz, E., et al. The psychology of word use in depression forums in English and in Spanish: testing two text analytic approaches. In: International Conference on Weblogs & Social Mediahttps://doi.org/10.1016/j.trstmh.2007.10.016 (2008)
32. Park, M., Mcdonald, D. W., Cha, M. Perception differences between the depressed and non-depressed users in twitter. In: International Conference on Weblogs and Social Media. (AAAI press, 2013)
33. Rude, S; Gortner, E; Pennebaker, J. Language use of depressed and depression-vulnerable college students. Cognit. Emot.; 2004; 18,
34. Molendijk, ML; Bamelis, L; van Emmerik, AA; Arntz, A; Haringsma, R; Spinhoven, P. Word use of outpatients with a personality disorder and concurrent or previous major depressive disorder. Behav. Res. Ther.; 2010; 48,
35. Lorenz, TA; Meston, CM. Associations among childhood sexual abuse, language use, and adult sexual functioning and satisfaction. Child Abuse Neglect; 2012; 36,
36. Gao,R., Hao, B., Li, L., Bai, S., Zhu,T. Shao. The establishment of Chinese language psychological analysis software system. In: Psychology and innovation enhancement - Proceedings of the 16th National Psychology Conference, 364–366 (2013)
37. Wu, P; Zhao, N; Li, S; Liu, Z; Wang, Y; Liu, T; Liu, X; Zhu, T. Exploring the psychological effects of COVID-19 home confinement in China: a psycho-linguistic analysis on weibo data pool. Front. Psychol.; 2021; 12, 2087.
38. Borsboom, D; Cramer, AOJ. Network analysis: An integrative approach to the structure of psychopathology. Annu. Rev. Clin. Psychol.; 2013; 9, pp. 91-121. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23537483][DOI: https://dx.doi.org/10.1146/annurev-clinpsy-050212-185608]
39. van Borkulo, CD; van Bork, R; Boschloo, L; Kossakowski, JJ; Tio, P; Schoevers, RA; Waldorp, LJ. Comparing network structures on three aspects: A permutation test. Psychol. Methods; 2022; 2022, pp. 1-13.
40. Bringmann, LF; Elmer, T; Epskamp, S; Krause, RW; Schoch, D; Wichers, M; Snippe, E. What do centrality measures measure in psychological networks?. J. Abnorm. Psychol.; 2019; 128, pp. 892-903. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31318245][DOI: https://dx.doi.org/10.1037/abn0000446]
41. Friedman, J; Hastie, T; Tibshirani, R. Sparse inverse covariance estimation with the graphical LASSO. Biostatistics; 2008; 9,
42. Tausczik, YR; Pennebaker, JW. The Psychological Meaning of Words: LIWC and Computerized Text Analysis Methods; 2010; SAGE Publications:
43. Lai, L. The Prediction of Language Use and Affective Tendency on Counseling Process and Outcome in Psychological Counseling; 2019; Central China Normal University:
44. Tausczik, YR; Pennebaker, JW. The psychological meaning of words: LIWC and computerized text analysis methods. J. Language Soc. Psychol.; 2010; 29,
45. Pennebaker, J. W., Chung, C. K., Ireland, M., Gonzales, A. & Booth, R. J. The development and psychometric properties of LIWC2007. http://www.liwc.net/LIWC2007LanguageManual.pdf (2007)
46. Chen, X; Jiang, Y. Empathy fatigue in counselors: connotations, influences, and preventive factors. Psychol. Res.; 2022; 15,
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.