The decline in birth rates and the surrounding occupational uncertainty have placed Chinese pre-service teachers under considerable job Stress (JS). While such Stress is often linked to detrimental effects, research in positive psychology suggests that adversity may also serve as a catalyst for growth (Tedeschi and Calhoun, 2004). To explore this mechanism, the present study integrates three complementary perspectives: the Transactional Model of Stress and Coping (Lazarus and Folkman, 1984; Folkman, 2013) (hereafter referred to as the transactional model), the Conservation of Resources (COR) theory, and the Challenge–Hindrance Stressor Framework (LePine et al., 2005; Cavanaugh et al., 2000).

The transactional model provides a macro framework for understanding the relationship between JS and individual outcomes, emphasizing that Stress results from cognitive appraisals rather than direct external stimuli. In this process, primary appraisal involves evaluating JS in terms of threat, loss, or challenge—a distinction central to the Challenge–Hindrance Stressor Framework. Secondary appraisal highlights the role of psychological Resources in determining coping efficacy (Folkman and Moskowitz, 2004). Within this framework, psychological resilience (PR) and future time perspective (FTP) enable individuals to situate present Stress within a broader life trajectory, viewing it as a manageable challenge and a necessary step in career development (Robitschek, 1998; Kooij et al., 2018; Carstensen, 2006).

COR theory elucidates the motivational mechanisms underlying this process, positing that individuals are fundamentally motivated to acquire, maintain, and expand valued Resources. Stress arises when Resources are threatened or lost. According to the principle of resource primacy, individuals with abundant “resource caravans” are better equipped to initiate “gain spirals” through resource investment (Hobfoll, 2001). In this context, PR provides the emotional and cognitive strength to prevent resource loss from escalating, while FTP offers motivational meaning by framing current sacrifices as investments for future returns (Kooij et al., 2018; Fasbender et al., 2019). Personal growth initiative (PGI) represents a typical form of resource investment, where individuals actively generate new Resources to counterbalance Stress and promote positive spirals (Robitschek, 1998; Weigold et al., 2024).

In sum, the integration of these theories suggests that the transformation from JS to PGI is not automatic; it requires enabling psychological Resources. This framework offers a nuanced understanding of how pre-service teachers can leverage internal Resources—where PR provides the “capacity to act” and FTP provides the “reason to act”—to navigate structural disadvantage and foster proactive growth.

The participants of this study comprised pre-service teachers and recent graduates (within 2 years of graduation) from Chinese teacher-training colleges and normal universities. This specific population was selected because they have been exposed to heightened employment-related Stress in the context of China’s declining birth rate, which has precipitated kindergarten closures, teacher redeployment, and a contraction of teaching positions. Furthermore, this cohort actively monitors recruitment trends (2024–2025) via digital platforms such as Douyin and Xiaohongshu, ensuring timely and informed perceptions of the labor market. Crucially, their reliable self-reporting capacity for psychological resilience and future time perspective makes them highly suitable for testing the proposed theoretical model: “Job Stress → Psychological Resilience/Future Time Perspective → Personal Growth Initiative.”

To ensure sample validity, participants were required to meet the following inclusion criteria: (1) majoring in a teacher-training discipline; (2) being current students or graduates from the classes of 2024 or 2025; and (3) possessing a clear understanding of the contemporary employment landscape, including specific terms such as “teacher establishment” (bianzhi) and “job competition.”

Data were collected between July 11 and July 24, 2025, through purposive sampling via the “Wenjuanxing” online platform, with a primary focus on vocational teacher-training institutions. A multi-stage quality control mechanism was implemented to ensure data integrity: (1) a preliminary screening item at the start of the survey; (2) embedded attention-check questions to identify inattentive responses; and (3) back-end data cleaning to exclude responses with logical inconsistencies or implausibly short completion times.

Initially, 1,181 responses were collected. Following the screening process, 953 valid questionnaires were retained, yielding an effective response rate of 80.7%. According to the widely accepted guideline of a subject-to-item ratio of 5:1 to 10:1 (Wang and Ji, 2020), the 49 measurement items in this study (JS: 5; PR: 23; FTP: 5; PGI: 16) required a minimum sample size of 245 to 490. The final sample of 953 significantly exceeds this threshold, providing a robust foundation for subsequent analyses. As shown in Table 2, the demographic profile—predominantly female (76.6%), aged 19–22 (82.4%), and vocational college students (94.6%)—authentically reflects the characteristics of pre-service teachers in the current job market. The demographic data of the research sample (N = 953) is as follows (Table 2).

The ethical conduct of this study complied with the guidelines for non-clinical social science research at Ganzhou Teachers College. Given that the study employed a cross-sectional, non-invasive, and low-risk anonymous survey design, it was exempted from formal review by the Institutional Review Board (IRB) in accordance with institutional policies. Prior to data collection, the survey protocol was submitted to and approved by the university’s internal research administration platform.

Informed consent was obtained from all participants through an online consent procedure. At the beginning of the questionnaire, participants were presented with an introductory statement clearly explaining the study purpose, the voluntary nature of participation, the estimated time required to complete the survey (approximately 3–5 min), data usage, and confidentiality protections. The statement explicitly informed participants that the survey was conducted anonymously, that no personally identifiable information would be collected, and that all data would be used solely for academic research and reported only in aggregated form. Participants were also informed that they could discontinue the survey at any time without any negative consequences.

Participants were required to indicate their agreement to participate after reading the informed consent statement, and completion and submission of the questionnaire were construed as providing informed consent. All data were collected anonymously, stored securely, and analyzed exclusively for research purposes, ensuring full confidentiality of participant responses.

The measurement instruments were adapted from well-established scales to ensure theoretical consistency and contextual relevance. All constructs were operationalized as latent variables measured by multiple items, as detailed in Tables 1, 3. To fit the specific context of Chinese pre-service teachers, minor linguistic adaptations were made to the original items without altering their core theoretical meanings.

All items were assessed using a 5-point Likert scale, ranging from “1 = strongly agree” to “5 = strongly disagree.” Higher composite scores indicate higher levels of the respective constructs. The specific instruments and their psychometric properties in the current sample are described below.

The theoretical model was analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS 3.3.9 software (Ringle et al., 2015). This approach was selected as it is highly effective for identifying key driver constructs in exploratory research rather than mere model confirmation (Hair et al., 2011). Compared to covariance-based methods, PLS-SEM is a robust, less restrictive approach that focuses on the explained variance of the criterion variables (Hair et al., 2012; Barclay et al., 1995).

The analysis followed a two-step procedure. First, the measurement model was evaluated to ensure the reliability and validity of the adapted scales (see Table 1 for operational definitions and Table 3 for items). Second, the structural model was assessed to test the hypothesized relationships. Given that this study explores the psychological mechanisms of personal growth initiative under specific structural Stress, PLS-SEM is considered an appropriate tool to handle the potential non-normal distribution of survey data and the complexity of the mediating pathways (Hair et al., 2022).

The assessment of convergent validity and reliability for the 11 constructs is summarized in Table 4. Standardized factor loadings for all observed indicators ranged from 0.877 to 0.958, exceeding the 0.70 threshold. Internal consistency was confirmed by Cronbach’s α (0.945–0.972) and composite reliability (CR: 0.958–0.978) values. Furthermore, the average variance extracted (AVE) for all constructs ranged between 0.821 and 0.901, surpassing the 0.50 benchmark and indicating strong convergent validity. This study used the Fornell–Larcker criterion to examine the discriminant validity of the four core constructs (Fornell and Larcker, 1981). As shown in Table 5, the square roots of the AVE for Job Stress (JS), Psychological Resilience (PR), Future Time Perspective (FTP), and Personal Growth Initiative (PGI) were 0.906, 0.877, 0.949, and 0.966, respectively. Each value was greater than the highest correlation coefficient between the construct and any other latent variable, confirming that the variables are conceptually and statistically distinct.

Discriminant validity was assessed using the Fornell–Larcker criterion (Table 5) (Chen et al., 2021). For each construct, the square root of AVE (diagonal) exceeded its correlations with other constructs, indicating adequate discriminant validity.

Before testing path coefficients, we assessed multicollinearity following Hair et al. VIF values ranged from 2.14 to 8.35, with the highest values occurring for PR and FTP (which correlated at r = 0.801). While none exceeded the critical threshold of 10, several approached the conservative cut-off of 5, suggesting moderate collinearity. Given that (1) discriminant validity was established via Fornell–Larcker criterion, (2) the constructs are theoretically distinct, and (3) PLS-SEM is relatively robust to multicollinearity, we proceeded with the analysis while acknowledging this limitation.

Path coefficients, t-statistics, and p-values are presented in Table 6 and Figure 2. The model demonstrated strong overall fit (GOF = 0.575).

Counter to our predictions, H1 and H2 were not supported. Job Stress positively—rather than negatively—predicted both Future Time Perspective (β = 0.369, t = 10.919, p < 0.001) and Psychological Resilience (β = 0.404, t = 11.840, p < 0.001). This unexpected pattern suggests that employment Stress may be functioning as a challenge Stressor in this context, mobilizing rather than depleting psychological Resources. Possible explanations are explored in the Discussion section.

As hypothesized, H3 and H4 received support. Future Time Perspective (β = 0.642, t = 13.161, p < 0.001) and Psychological Resilience (β = 0.314, t = 6.412, p < 0.001) both significantly predicted Personal Growth Initiative.

This study contributes to the literature on occupational Stress and vocational development by interrogating the mechanism through which structural job Stress correlates with proactive growth.

First, the findings extend the Challenge–Hindrance Stressor Framework by situating it within the structural labor market contraction facing emerging adults. Contrary to the conventional COR assumption that employment threats are inherently resource-depleting, the current results suggest that JS may be positively associated with psychological Resources under specific developmental conditions. This aligns with a “gain-spiral” interpretation, suggesting that when Stressors are appraised as bounded, they may serve as a mobilizing condition for short-term resource investment rather than immediate exhaustion.

Second, the study refines the Stress–growth pathway by identifying the relative salience of cognitive–motivational Resources. Although PR and FTP are often conceptualized as parallel mediators, FTP demonstrates a substantially stronger predictive relevance for PGI. This finding resonates with SST, suggesting that during early career transitions, a sense of future openness acts as a more proximal predictor of intentional self-development than affective coping (resilience) alone.

Finally, by focusing on vocational pre-service teachers—a group often framed through a “deficit lens”—this research challenges vulnerability-based frameworks. The evidence suggests that students in structurally constrained contexts may nevertheless maintain agentic orientations, thereby broadening the discourse on teacher development to include proactive resilience within non-elite educational pathways.

The findings offer strategic insights for teacher education programs adapting to China’s shifting demographic and employment landscape.

Temporal Interventions: Given the role of FTP, institutions should move beyond placement services to integrate structured future-oriented exercises. Possible-selves workshops and diversified career mapping can help students construct broader temporal horizons, mitigating the anxiety associated with a single-track bianzhi (tenure) narrative.

PsyCap ecosystems: Rather than relying on reactive counseling, programs might benefit from embedding PR training and growth-mindset modules directly into the curriculum. Combining these with experiential learning (e.g., simulated job-search scenarios) can provide a form of “Stress inoculation,” supporting adaptive engagement with professional uncertainty.

Protean career narratives: Universities and policymakers should consider reframing employment Stress as a normative developmental hurdle. Promoting a “protean career” lens—emphasizing transferable skills and lifelong adaptability via new media and alumni role models—can encourage students to interpret structural constraints as opportunities for agentic growth.

Several constraints must be noted when interpreting these findings.

Most importantly, the cross-sectional nature of the data precludes definitive causal claims. While PLS-SEM identifies directional paths, the results represent statistical associations rather than confirmed effects. Reciprocal relationships—whereby individuals with high PGI are more predisposed to appraise JS as a challenge—cannot be ruled out. Future research should employ longitudinal designs, such as cross-lagged panel models, to verify the temporal ordering and the long-term sustainability of the observed resource mobilization.

Furthermore, the reliance on self-reports raises the potential for Common Method Variance (CMV). The robust correlation between FTP and PGI, while statistically meeting discriminant validity criteria, hints at conceptual proximity. Subsequent studies would benefit from incorporating multi-source data (e.g., peer or supervisor ratings) or behavioral indicators to enhance measurement objectivity.

Lastly, the sample homogeneity—predominantly vocational students from a single region—limits the generalizability of the results. Replicating this model across diverse educational tiers and gender-balanced cohorts is essential to determine whether this “Stress-growth” mechanism is universal or context-specific. Additionally, exploring nonlinear pathways would be valuable to identify the threshold at which employment-related Stress shifts from a growth-catalyst to a resource-hindrance.