Collected data included gender, age, educational background, professional title, employment type, hospital type, years of work experience, whether the participant is an only child, marital status, average monthly income, whether the participant is a specialized nurse, department, involvement in department management, participation in teaching tasks within the department, and gender of the head nurse. A total of 15 items were included.

The MWMS, refined by Gagné et al. (2015), is an assessment tool based on the multidimensional concept of motivation within Self-Determination Theory (SDT). It is used to measure types of employee work motivation. This scale has been validated across multiple countries and languages, making it suitable for assessing work motivation in cross-cultural contexts. It comprises six dimensions: Amotivation, External Regulation—Social, External Regulation—Material, Intrinsic Motivation, Identified Regulation, and Introjected Regulation, with a total of 19 items. A Likert 7-point rating scale (ranging from 1 to 7, corresponding to “completely disagree” to “completely agree”) is employed. The total score ranges from 19 to 133, with higher scores indicating stronger motivation toward a specific type. The Cronbach’s α coefficient of this scale demonstrates high internal consistency across different subscales and language versions. In the original study, most subscales had Cronbach’s α coefficients >0.70, and in this study, the Cronbach’s α coefficient for the overall scale was 0.928, indicating high internal consistency.

The Voice Behavior Scale, designed by Liang et al. (2012), is an assessment tool aimed at measuring employees’ levels of voice behavior within an organization. It consists of two dimensions: proactive voice behavior and prohibitive voice behavior, with a total of 10 items. Each item uses a Likert 5-point rating scale, ranging from “strongly disagree” to “strongly agree,” corresponding to scores of 1 to 5, respectively. The total score of the scale ranges from 10 to 50, with higher scores reflecting more positive voice behavior among employees. The Cronbach’s α coefficient of this scale is 0.915, indicating good internal consistency. In this study, the Cronbach’s α coefficient for this scale was 0.962.

Prior to the commencement of the study, a pretest was conducted among 30 randomly selected participants who met the inclusion and exclusion criteria from the aforementioned six hospitals. During the pretest, the research team closely monitored participants’ feedback and recorded the time taken to complete the questionnaire, as well as any questions or suggestions they had regarding various sections of the questionnaire. Based on their suggestions and confusion, revisions and improvements were made to the questionnaire. In the pretest, the Cronbach’s α coefficient for the Nurse Voice Behavior Scale was 0.929, and the Cronbach’s α coefficient for the Multidimensional Work Motivation Scale (MWMS) was 0.774.

The data collected through an online survey platform¹ were exported to an Excel spreadsheet and verified by two researchers independently. Statistical analysis was conducted using SPSS 26.0 software. For continuous variables, the mean ± standard deviation was used for description, while for categorical variables, frequency, composition ratio, and rates were employed. Chi-square tests or rank-sum tests were applied for comparisons between groups. Analysis of Variance (ANOVA) was used for normally distributed data, while non-parametric tests were utilized for skewed distributions.

Confirmatory factor analysis (CFA) of the MWMS was conducted using Mplus 8.3 software. Competitive models, including a six-factor model, a five-factor model, and a three-factor model, were constructed to assess the goodness-of-fit. The overall fit of the models was reflected by the ratio of the chi-square value (Χ²) to the degrees of freedom (df). The Comparative Fit Index (CFI) and Tucker–Lewis Index (TLI), serving as incremental fit indices, indicated better model fit as their values approached 1. A smaller Standardized Root Mean Square Residual (SRMR) was preferred, generally considered good if less than 0.08. Similarly, a smaller Root Mean Square Error of Approximation (RMSEA) was better, with values between 0.05 and 0.08 indicating a good fit and values less than 0.05 indicating an excellent fit.

Latent Profile Analysis (LPA) of nurses’ voice behavior was conducted in Mplus 8.3, starting with a baseline model of a single class and gradually increasing the number of classes. Model selection was based on the Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC), and adjusted Bayesian Information Criteria (aBIC), with lower values indicating better model fit. An entropy value closer to 1 indicated higher classification accuracy. The Lo–Mendell–Rubin (LMR) test and the Bootstrap Likelihood Ratio Test (BLRT) were used to assess differences in fit between models, with a p-value less than 0.05 indicating that a model with an increased number of classes was significantly better than the previous one. When determining the optimal model, the practical significance of the classification was also considered.

Using the classification results from LPA as the dependent variable, multivariate Logistic Regression Analysis was conducted on variables that were statistically significant in the univariate analysis.

Given that all data in this study were self-reported and collected through the same online platform “Wenjuanxing,” there is a potential risk of common method bias (CMB). To assess this potential concern, we performed Harman’s single-factor test. The results showed that the first factor accounted for 51.272% of the total variance, which is well below the 60% threshold recommended by Podsakoff et al. (2003). This indicates that common method bias is not a serious issue in the present study.

Within the scope of this study, the core characteristics of “Conservative” nurses were young (<30 years old), junior professional title, non-establishment (Unstable employment status), working experience < 10 years, unmarried, not involved in department management and not undertaking teaching tasks (Table 1). This result can be interpreted through three interconnected lenses: (1) Motivational suppression: Amotivation directly undermines the internal drive for voice behavior, aligning with SDT’s core principle that amotivation inhibits proactive behaviors (Deci and Ryan, 2008). (2) Implicit voice theories: Consistent with Şahin et al. (2021), Conservative nurses strongly endorsed beliefs such as “negative career consequences of voice” and “don’t embarrass the boss in public”, which foster defensive silence to avoid perceived risks. In Guangxi’s hierarchical medical environment, junior nurses may internalize these beliefs to maintain occupational safety, as IVTs often operate as taken-for-granted norms in collectivistic cultures (Şahin et al., 2021). (3) Occupational insecurity: Non-tenured status and limited work experience reduce perceived job security, reinforcing a risk-averse attitude toward voice (Li and Song, 2024).

The “Balanced Risk-Taker” profile shares similar demographic characteristics with the “Conservative” profile (i.e., younger age, junior professional title, non-tenured status, less than 15 years of work experience, unmarried, and no involvement in management or teaching responsibilities) (Table 1). However, regression results revealed that these nurses were more distinguishable from the “Conservative” (C1) profile due to their simultaneously higher scores in both amotivation and intrinsic motivation, along with being unmarried. This “motivational conflict” reflects the professional dilemma faced by early-career nurses: on the one hand, intrinsic motivation fosters a degree of enthusiasm and willingness to improve their work; on the other hand, amotivation (potentially stemming from occupational stress or uncertainty about career development) suppresses their initiative to speak up, ultimately resulting in a behavioral pattern characterized as “weighing risks and benefits while voicing opinions prudently.” Compared to the “Conservative” profile, these nurses possess slightly more work experience and a better capacity for risk assessment, enabling them to seek a balance between “maintaining the “advocating for minor changes.” The existence of this subgroup further demonstrates the heterogeneity in nurses’ voice behavior, suggesting that traditional binary classification approaches are inadequate for capturing such intermediate profiles.

In this study, “Transitional” nurses accounted for the highest proportion (34.38%), with core characteristics including older age (>30 years old), intermediate professional title, permanent employment status within China’s healthcare system (i.e., stable and formal employment), work experience exceeding 5 years, and assuming preceptorship responsibilities without participating in departmental management (Table 1). Regression results indicated that these nurses were more likely to transition from “conservative-type” to “active-type” due to higher scores in external regulation-material, identified regulation, and intrinsic motivation, as well as undertaking preceptorship tasks and assisting in departmental management. This finding aligns with Karataş and Çankır (2023), who emphasized that collegial solidarity and administrative support—key contextual factors for transitional nurses—mediate the relationship between individual motivation and work-related proactive behaviors. Specifically, the preceptorship role endows them with the identity of knowledge transmitters, prompting attention to nursing process optimization (Chen et al., 2012). Additionally, the mediating role of thriving at work (a construct integrating learning and vitality; Porath et al., 2012) may further explain this transition, as autonomous motivation fosters thriving, which in turn promotes constructive voice (Ozcan et al., 2023; Karataş and Çankır, 2023). Meanwhile, the long-term career stability brought by permanent employment status (i.e., “perceived job security from stable employment” in international research) further reduces the perceived risk of voice behavior (Falatah et al., 2021). Combined with the regional cultural characteristics of Guangxi, these nurses, influenced by collectivist culture, emphasize team stability and also possess a certain willingness for change due to professional maturity, thus forming the voice characteristic of “steady transition.” This finding also suggests that in collectivist-oriented healthcare settings, “intermediate-type” voice actors may serve as key bridges for promoting organizational change.

The “Challenging” profile represents the most change-oriented group of nurses, characterized by seniority (over 15 years of work experience), intermediate or senior professional titles, tenured status, being married, and involvement in both departmental management and clinical teaching. Regression results further clarified that nurses with higher scores in identified regulation and intrinsic motivation, and those involved in departmental management, were more likely to belong to this profile. When compared to the “Transitional” (C3) profile, “Challenging” nurses exhibited significantly higher levels of both identified regulation and intrinsic motivation. This finding strongly supports the core assumption of Self-Determination Theory (SDT) that autonomous motivation, which encompasses both intrinsic motivation and identified regulation, serves as a key driver prompting individuals to engage in proactive voice behavior (Ryan and Deci, 2020; Chen et al., 2024). The professional confidence gained from seniority, the sense of security associated with tenured positions, and the voice afforded by participation in management collectively reinforce their willingness to challenge the status quo through their voice behavior. From a managerial perspective, the presence of such nurses injects innovative vitality into organizations, and their development is closely linked to the presence of a supportive environment for voice behavior within the organization, exemplified by access to management participation channels and opportunities for career development.

Autonomous motivation—comprising intrinsic motivation and identified regulation—emerged as the most potent predictor of active voice behavior. Nurses with higher scores in these dimensions were significantly more likely to belong to the Transitional (C3) and Challenging (C4) profiles. The profound predictive power of identified regulation for the Challenging profile underscores that when nurses internalize the value of their work, voice behavior transforms from discretionary participation to a professional obligation aligned with their self-concept (Ryan and Deci, 2020). This mechanism is consistent with Ozcan et al. (2023), who found that autonomous motivation (especially identified regulation) enhances thriving by satisfying psychological needs for autonomy and competence, thereby fostering sustained proactive behaviors. In the context of nursing, challenging-type nurses’ high identified regulation reflects their alignment of voice behavior with professional ethics and patient safety goals, while their high intrinsic motivation fuels continuous learning and innovation in care delivery (Ozcan et al., 2023). Together, these motivational processes create a virtuous cycle: autonomous motivation promotes thriving, which in turn strengthens constructive voice (Karataş and Çankır, 2023). This aligns with SDT’s tenet that autonomous motivation fulfills psychological needs for autonomy and competence, thereby fostering proactive and sustained behaviors (Gagné et al., 2015). However, the cross-sectional design necessitates caution against causality; longitudinal studies are needed to explore whether autonomous motivation drives voice or vice versa.

To effectively enhance the sense of autonomy, the strategies we can adopt include: providing diversified career development opportunities (Kohnen et al., 2023), increasing the interest and challenge of work content (Bogdán et al., 2024), and creating a positive and supportive work environment (Aleo et al., 2024). These targeted initiatives are expected to motivate nurses to more proactively engage in nursing work, thereby driving an overall improvement in nursing quality and work efficiency (Zeng et al., 2022).

Amotivation consistently predicted membership in conservative profiles (C1 and C2). Its positive association with the Balanced Risk-Taker profile (C2)—despite this group’s motivation—reveals a “motivational conflict.” This paradox reflects early-career nurses’ dilemma: intrinsic interest is stifled by uncertainties in career development or perceived organizational indifference, leading to a cautious, risk-weighing approach to voice (Li and Song, 2024). For Conservative nurses (C1), amotivation directly undermines the internal drive for voice, aligning with SDT’s emphasis on amotivation as a barrier to proactive behavior (Deci and Ryan, 2008).

However, in interpreting the motivational conflict specific to the Balanced Risk-Taker profile (C2, n = 38, 5.42%), caution is warranted due to its relatively smaller sample size compared to other profiles. While the observed association is statistically significant, its magnitude and generalizability should be viewed as preliminary. Future research with larger, balanced samples is needed to confirm the robustness of this nuanced motivational dynamic.

To address amotivation, establishing a fair and transparent evaluation mechanism is crucial (Li et al., 2021). Such a system should ensure rewards accurately reflect contributions, building trust in the reward-punishment framework (Ryan and Deci, 2020) and mitigating the inhibitive effects of amotivation.

The material dimension of extrinsic regulation showed a significant positive association with active voice behavior, confirming that tangible rewards can reduce the perceived “risk cost” of voice, especially among non-tenured nurses (Fedele, 2018). However, relevant research (Fischer et al., 2019) demonstrates that long-term excessive reliance on material rewards and punishments may undermine nurses’ intrinsic motivation, leading to the so-called “crowding-out effect”—where extrinsic incentives weaken the intrinsic drive (e.g., passion for improving patient care) that originally fuels proactive voice. In this study, despite the positive association of the material dimension with voice behavior, the direct positive impact of social rewards and punishments (e.g., public recognition, career advancement) was not significant. This is likely due to contextual factors (e.g., junior nurses’ deference to senior staff in Guangxi’s hierarchical workplace norms, which reduces the motivational impact of public praise) and methodological limitations. Nevertheless, theoretically, social rewards and punishments are crucial for satisfying nurses’ needs for social belonging and career development, and thus should have a potential positive impact on voice behavior (Lee et al., 2022). Additionally, it is noteworthy that external pressure from controlled motivation (e.g., over-reliance on material rewards as a mandatory “incentive” rather than a supplementary one) may inhibit the sustainability and depth of voice behavior, implying that managers need to carefully consider the comprehensive impact of reward and punishment measures on nurses’ long-term work motivation and voice behavior when adopting them.

When formulating reward and punishment measures, nursing managers need to carefully weigh the proportion of material rewards and social rewards and punishments (Feng et al., 2024). Although material rewards may significantly boost nurses’ work enthusiasm in the short term (Fedele, 2018), long-term excessive reliance on such rewards may weaken nurses’ intrinsic motivation, affecting their intrinsic love and responsibility for their work (Deci and Ryan, 2000). Conversely, social rewards and punishments, such as public praise and career advancement, can satisfy nurses’ higher-level needs for social belonging, professional identity, and career development, thereby more effectively stimulating their voice behavior and work engagement (Ayalew et al., 2021; Liang et al., 2012).

Notably, in our nurse sample, the material and social dimensions of extrinsic regulation exhibited significant independence and variability—reflecting diverse motivational needs: some nurses (e.g., non-tenured staff) prioritize material stability, while others (e.g., mid-to-senior title nurses) value social satisfaction more (Abu Yahya et al., 2019). This diversity necessitates a deep understanding of individual needs to design precise incentives—for example, pairing targeted material rewards with low-pressure social recognition (e.g., private feedback) for non-tenured nurses, while prioritizing career advancement-linked social rewards for senior nurses.

The moderate correlation between introjected regulation and extrinsic-social regulation (r = 0.670) did not raise multicollinearity concerns (VIF < 10), yet neither dimension demonstrated significant predictive power. This can be explained from two perspectives: first, the externally pressured nature of introjected regulation (e.g., voicing to avoid criticism) is unlikely to sustain the initiative and depth of voice behavior (Deci and Ryan, 2000); second, the effectiveness of extrinsic-social regulation may be constrained by workplace hierarchy norms (e.g., junior nurses’ deference to senior staff could diminish the motivational impact of public recognition) (Zhang et al., 2024). These findings suggest that future management strategies should design low-pressure social reward formats, such as anonymous team appreciation or cross-hierarchical communication meetings.

The non-significant relationship between certain motivational factors and voice behavior profiles may be explained by the organizational context. Specifically, in environments with high power distance, such as healthcare settings in hierarchical cultures, the effect of individual motivation may be attenuated by organizational factors. This aligns with Şahin et al. (2021), who found that power distance negatively influences voice behavior, suggesting that organizational hierarchy can override individual motivational factors.

Conservative nurses face core barriers of occupational insecurity and hierarchy-induced silence, and strategies prioritize transformational leadership’s “individualized consideration” to address these pain points while leveraging perceived organizational support (Xia et al., 2023). Given that intrinsic motivation is crucial for stimulating work enthusiasm (a key finding from our analysis), nursing managers should provide targeted career development opportunities (e.g., skill training linked to promotion) and modest work autonomy (e.g., letting nurses schedule their daily patient follow-up tasks) to comprehensively stimulate intrinsic motivation—addressing the root of amotivation in this group. To mitigate stress-driven amotivation further, offering confidential one-on-one counseling and monthly stress-relief workshops (Pu et al., 2024) supports nurses’ well-being and helps them convert latent concerns into constructive voice. For hierarchy-related silence, launching an anonymous digital feedback platform (e.g., a WeChat mini-program) themed around “team safety” (e.g., “suggestions to improve patient handoff processes”)—complemented by physical “nurse suggestion boxes” in ward areas—frames voice as a “collective responsibility,” aligning with local collectivist values and mitigating junior nurses’ deference to senior staff (Han and Xu, 2024); this multi-channel approach strengthens their sense of participation while reducing the perceived “risk cost” of speaking up. Simultaneously, establishing a positive feedback system (e.g., weekly team meetings acknowledging even small constructive suggestions) ensures timely and fair recognition of contributions, enhancing job satisfaction and organizational belonging (Feng et al., 2024; Xia et al., 2024).

Balanced Risk-Takers require low-risk interventions to convert intrinsic drive into actionable voice. Since intrinsic motivation and identified regulation jointly drive work enthusiasm, managers can increase the interest and challenge of work content (e.g., assigning them to lead small “care process optimization” sub-tasks) to nurture intrinsic motivation, while explaining how these tasks align with nursing professionalism to reinforce identified regulation. Establishing a transparent, merit-based reward mechanism—where rewards tie to voice quality (e.g., “adopted suggestions that enhance care efficiency earn 5% performance bonuses”) rather than quantity—ensures rewards accurately reflect nurses’ actual contributions (Li et al., 2021), building trust in the reward-punishment system (Ryan and Deci, 2020) and reducing amotivation from “unfair treatment.” Complementing this with low-pressure social recognition (e.g., one-on-one acknowledgment from supervisors instead of public praise) avoids triggering discomfort from hierarchical norms while satisfying their need for social validation (Lee et al., 2024); pairing this recognition with a positive feedback system (e.g., written notes of appreciation for thoughtful suggestions) further strengthens organizational belonging, helping resolve their motivational conflict.

Transitional nurses act as “change bridges” between conservative and active voice behaviors. Since identified regulation motivates nurses to engage in professional behaviors aligned with self-concept, managers should encourage this group to participate in department management and decision-making (e.g., inviting them to join monthly “ward improvement meetings”) to enhance their autonomy, sense of responsibility, and professional identity; this aligns with our finding that identified regulation strengthens proactive voice. Developing a “Preceptor Voice Workshop”, where nurses teach trainees to both identify care gaps and propose feasible solutions (e.g., “how to suggest workflow tweaks during clinical teaching”), further ties their preceptorship role to proactive voice, reinforcing professional identity and deepening identified regulation (Chen et al., 2012). Department managers can model constructive voice by sharing “voice success cases”(e.g., “a Transitional nurse’s suggestion reduced catheter-related infections by 20%”) in team meetings—framing voice as a marker of “professional maturity” to align with their focus on team stability (Balay-Odao et al., 2022)—and complement this with regular nurse forums to provide structured platforms for opinion exchange, strengthening their sense of participation and belonging (Schaufeli and Bakker, 2004). Additionally, designating “team optimization forums” aligns with collectivist culture, positioning voice as a shared responsibility to reduce resistance to change while sustaining their “steady transition” behavior.

Challenging nurses serve as organizational innovators, and strategies prioritize transformational leadership’s “inspirational motivation” to maintain their drive, integrating insights into intrinsic motivation and identified regulation. As intrinsic motivation not only stimulates work enthusiasm but also promotes innovative problem-solving, managers should provide diversified career development opportunities (e.g., fellowships in nursing innovation, advanced clinical training) and greater work autonomy (e.g., letting them design new patient care protocols) to sustain intrinsic motivation—supporting their adoption of innovative methods to address work challenges (Wei et al., 2019). Encouraging deeper participation in department management and decision-making (e.g., appointing them to core nursing quality committees, such as “care innovation steering groups”) aligns with their identified regulation motivation, further enhancing professional identity and reinforcing proactive voice. Granting leadership roles in cross-departmental projects (e.g., piloting smart nursing tools) leverages their experience to spread proactive voice norms, while establishing a “nursing innovation archive”—where impactful suggestions (e.g., those reducing patient readmission rates) are formally documented and linked to career advancement (e.g., promotion prioritization)—validates their role as “change leaders.” Regular nurse forums and digital feedback platforms (in addition to their management roles) provide additional channels for them to express opinions, and a positive feedback system (e.g., organizational awards for “most impactful voice contributions”) ensures timely recognition of their innovations, sustaining autonomous motivation and organizational belonging (Feng et al., 2024; Schaufeli and Bakker, 2004).