The prevalence of depressive symptoms among current students is becoming more concerning, frequently resulting in severe repercussions for their well-being. The relationship between depressive symptoms and students' well-being is a critical area of research, particularly highlighted during the COVID-19 pandemic. Studies indicate that depressive symptoms significantly correlate with various aspects of well-being, including life satisfaction, physical health, and overall psychological health (Ochnik et al., 2024). A moderate positive connection exists between depressive symptoms and psychological well-being in teenagers (Alshehri et al., 2023). Life satisfaction and physical health were recognized as protective factors against depressive symptoms, highlighting the necessity for therapies that improve these dimensions (Ochnik et al., 2024). Variables such as family income and parental occupation were significantly associated with depressive symptoms, indicating that socio-economic status plays a role in students' well-being (Alshehri et al., 2023). While existing research establishes a significant correlation between depressive symptoms and students' well-being – encompassing factors like life satisfaction, physical health, overall psychological health, and socioeconomic status (family income, parental occupation) – there remains a critical gap. Specifically, the relationship between depressive symptoms and well-being understood through the lens of micro psychological activities (such as psychological resilience and egoism) is inadequately explored. This gap limits our understanding of how these fundamental cognitive and emotional processes contribute to, or protect against, depressive symptoms in students, particularly when considered alongside established socioeconomic and broader well-being factors.

This study examined the relationship between depressive symptoms and students' well-being, with psychological resilience serving as a mediator and egoism as a moderator. Psychological resilience has been shown to reduce the impact of depressive symptoms on well-being. For instance, higher resilience levels correlate with improved quality of life in individuals with depression, suggesting that resilience can shield from the adverse effects of depressive symptoms (Novak and Lev-Ari, 2023). Studies indicate that resilience traits, such as hardiness, can lessen the negative correlation between depression and well-being, allowing individuals to maintain a fulfilling life despite depressive symptoms (Chuning et al., 2024). Egoism can mitigate the connection between depression and well-being by affecting emotional regulation and coping mechanisms. Elevated egoism (positive self-concept) and sophisticated defense mechanisms correlate with enhanced well-being and reduced depression (Padmanabhanunni and Pretorius, 2022).

Previous research exclusively utilized SmartPLS to testify the relation between depressive symptom and students' well-being; this study employs a combination of fuzzy Delphi theory and SmartPLS. The integration of FDM and SmartPLS can yield more robust educational models by amalgamating qualitative expert insights with quantitative data analysis (Suib et al., 2024). The integration of FDM and SmartPLS can provide more robust educational models by amalgamating qualitative expert insights with quantitative data analysis. This combination improves the capacity to obtain expert consensus while concurrently examining intricate correlations among educational variables (Suib et al., 2024).

Interventions addressing depressive symptoms across core domains (emotional, cognitive, somatic, behavioral) are consistently associated with enhanced student well-being, supported by a significant positive correlation with psychological well-being (Park et al., 2023). Heightened psychological resilience within specific subdimensions highlights the critical role of structured social support systems in alleviating stress within socially demanding environments (Çaglar et al., 2025). Psychological resilience acts as a vital protective influencing factors, mitigating symptom impact through enhanced emotional regulation and adaptive coping, particularly within demanding academic settings (Çaglar et al., 2025). Furthermore, egoism significantly moderates the link between depressive vulnerability and interpersonal challenges, suggesting a stable self-concept aids in navigating stressors without succumbing to depression (Han and Lee, 2019). Seo (2022) confirms egoism's protective function through its alleviation of school stress and depression. Therefore, this study aims to examine the interrelationships between core depressive dimensions (emotion, cognition, somatic, behavioral), psychological resilience, and egoism in shaping college students' mental health and well-being.

Specifically, based on this empirical foundation and the proposed research purpose, we propose the following hypotheses:

In conclusion, this research significantly advances our understanding by moving beyond a simple link between depression and student well-being. It illuminates the crucial mediating pathway of psychological resilience across specific depressive dimensions and identifies egoism as a key moderator. These insights are vital for developing targeted, mechanism-driven, and personalized interventions to effectively promote resilience and enhance well-being in diverse student populations facing depressive challenges. This framework delivers actionable pathways for educational institutions to translate theoretical insights into structured support systems, directly addressing the multifaceted nature of student depressive syndrome through evidence-based academic practice.

The Delphi approach, including the Delphi procedure and fuzzy theory, was employed to examine the perspectives of professionals knowledgeable in this field of study. This approach has been employed to attain anonymity, iteration, regulated feedback, and statistical group response. The implementation of the fuzzy Delphi technique necessitates a panel of experts including between nine and fifteen individuals (Daud et al., 2024). Consequently, this study focused on 10 experts with at least 5 years' expertise. This arises from the divergence of perspectives on this subject. For this study, 10 experts from China's vocational education were recruited. Table 1 displayed the particular circumstances of the experts. These 29 variables influencing student well-being were derived from a 2003–2023 meta-analysis (Xiong et al., 2025). They were structured into an expert's scale to collect expert opinions. Ten experts evaluated each variable using this instrument. The opinions of these professionals were solicited, as they had a responsibility to provide guidance about the improvement of college students' well-being. The questions utilized a 5-point Likert scale, with the five levels designated as “completely disagree,” “disagree,” “neutral,” “agree,” and “completely agree,” assigned corresponding reverse integers from 1 to 5. This method has six steps, beginning with experts assessing the significance of the variable, followed by the conversion of linguistic variables into fuzzy numbers, and culminating in the use of the vertex method to calculate the distance between the average values.

Table 1 delineated the characteristics of the experts and their replies to questions concerning the factors that affected students' well-being. The ages of the experts, as shown in Table 1, ranged from 21 to 50 years. They occupied various roles, including section chief, division chief, faculty member, director, principal, and instructor. The professionals' expertise ranged from 5–20 years. This study sought to determine the views of seasoned administrative personnel and proficient educators regarding the elements that impacted students' well-being. To ensure robust dimension validation grounded in both scholarly rigor and practical relevance, experts were purposively selected based on three integrated criteria: (a) a minimum 5 years' specialized experience in clinical psychology, educational psychology, or adolescent mental health intervention; (b) demonstrated scholarly contributions through publications in depression, resilience, or ego psychology; and (c) active engagement in university counseling or student well-being programming. This triangulation of credentials ensured panelists possessed the necessary theoretical-practical nexus to effectively evaluate student well-being dimensions—particularly micro-psychological constructs like egoism—within authentic educational ecosystems (Stern et al., 2020).

In the fourth phase, according to Khaira and Ibrahim (2024), if the disparity between the average and the expert evaluations is below the threshold of 0.2, all experts are deemed to have reached an agreement.

If the proportion of reaching group consensus among the m x n ratings of options and n criteria weights exceeds 75% (Khaira and Ibrahim, 2024), proceed to step 5; otherwise, a second round of survey is necessary.

The fifth step encompasses imprecise evaluations:

The six-step process for each of the available alternatives, the fuzzy evaluation Ã_i = (a_i1, a_i2, a_i3) was defuzzified by

The fuzzy Delphi Method aimed to achieve expert consensus with a questionnaire utilizing the Likert scale. A fuzzy Delphi method was employed to transform the results into fuzzy integers. The language variables were converted into fuzzy scales. The mean group consensus was 80%, rendering this version acceptable. Consensus would be achieved if the average group agreement exceeded 75%. If the percentage is below 75%, a new poll must be administered due to the absence of consensus (Buryanov et al., 2023). The principal acceptable and rejected dimensions, as determined by fuzzy findings, were illustrated by the number below (Figure 2).

Figure 2 indicated that the d values for emotional support, cognitive and mental health, somatic health, behavioral habits, extracurricular activities, and academic support were below 0.2. The six components pertained to the dimension of depressive symptoms. The values of psychological egoism, self-worth, and self-recognition that constituted the egoism dimension were below 0.2 (Peng et al., 2024). The value of psychological resilience was below the threshold of 0.2. Ultimately, three dimensions—depressive symptoms, psychological resilience, and egoism—were identified as the primary determinants impacting the enhancement of students' well-being.

This study selected one public institutions in Shandong Province that offered Traditional Chinese Medicine (TCM) programs as the research subjects. A mixed-methods strategy, integrating Fuzzy Delphi method and SmartPLS, was adopted. The sequential mixed-methods design was theoretically anchored in complexity theory, where qualitative expert consensus (Phase 1) captured context-bound conceptualizations of dynamic mental health constructs (depressive symptoms, egoism, resilience), enabling subsequent quantitative modeling (Phase 2) to test generalized relationships while preserving ecological validity (Stern et al., 2020). The quantitative phase involves modeling the relationships identified in the qualitative phase to test their generalizability across larger populations. This phase uses statistical methods to validate the findings and explore the broader applicability of the insights gained (Wackerhagen et al., 2022). Initially, structured expert interviews were performed to examine the elements affecting the students' well-being. The expert questionnaire was designed based on the research objective and the previous literature review, encompassing 29 items. After the analysis of expert questionnaire, three influencing factors related to students' well-being were decided such as depressive symptoms, egoism and psychological resilience. The “Multidimensional Assessment Scale of Mental Health for College Students” was developed based on the findings from the expert questionnaire analysis. The scale comprised four dimensions. The perspective on depression symptoms encompassed four aspects and 20 items derived from the UPI scale. The University Personality Inventory demonstrated exceptional internal consistency, achieving an overall alpha of 0.97 and individual variable alphas ranging from 0.83 to 0.91, indicating its dependability as a mental health screening tool (Sugawara et al., 2020). Five items pertaining to psychological resilience were derived from the psychological resilience scale. Psychological resilience scales are essential tools for evaluating an individual's ability to adapt and thrive during adversity (Reinwarth et al., 2024). Five items were selected from the psychological egoism scale to assess the concept of egoism. Psychological egoism scales are tools designed to evaluate the extent to which individuals prioritize self-interest in their thoughts and behaviors (Overall and Gedeon, 2022). Ten indicators of student well-being were derived from the school well-being profile (Monroy et al., 2024). The well-being dimension was divided into two components: academic well-being and emotional well-being. The “Multidimensional Assessment Scale of Mental Health for College Students” totally included 40 items.

This research was carried out in higher education institutions in China. Consequently, the research was conducted in the first quarter of 2025. Ten experts of this study were selected from the target population by purposive sampling procedures. Purposive sampling aims to select individuals possessing particular features or experiences pertinent to the study inquiry (Magnone and Yezierski, 2024). For the second phase (quantitative methodology), simple random sampling methodology was adopted. For quantitative part, simple random sampling methodology was adopted. In SRS, each unit in the population has an equal probability of being selected. This can be achieved by drawing units one by one, either with or without replacement, from the population (West, 2016). A total of 1382 students participated in this research, with the majority being female and the remainder male. All participants were aged 18 and above, excluding minors. In total, 1,336 students completed the questionnaires, resulting in an 97% response rate, which met the criteria of cross-sectional studies. Researchers intending to employ Structural Equation Modeling (SEM) must ensure a minimum of 10 participants per estimated parameter. Given 40 parameters, this necessitates at least 400 subjects (Shinya and Takiyama, 2024). Consequently, the sample size exceeding 400 cases was adequate for deriving significant results.

This research endeavor commenced with the assessment and endorsement of the Human Research Ethics Committee. Subsequently, the pertinent documents were gathered and examined to construct an expert questionnaire. The questionnaire was dispatched to 10 experts via email to elicit their responses regarding the significant aspects impacting students' well-being. Multidimensional Assessment Scale of Mental Health for College Students was developed and finalized based on expert analysis results. Upon the completion of the initial edition of the scale, a pilot study was executed utilizing the instrument with a sample from various majors of students, facilitating the refinement of the questionnaire for its final iteration. The students were subsequently invited to participate in this research voluntarily. The participants received the informed consent statement, which outlined the research objectives, duration, instrument instructions, data confidentiality assurances, and expected outcomes. Participants who executed this informed consent form consented to engage in the study on-site by completing the questionnaire. Individuals unable to attend in person answered the survey through an online link.

The rights of the subjects were consistently upheld throughout the investigation. Participants were informed that they could withdraw at any moment. The participants' physical and mental well-being, data confidentiality, the right to anonymity, and other ethical requirements pertaining to human subjects were safeguarded. Figure 3 illustrated the progression of the study.

The Partial Least Squares (PLS) modeling was employed to assess the study framework utilizing SmartPLS v4.0, as it analyzed data independent of sample size and normality. Moreover, SmartPLS possesses enhanced capabilities for predicting and analyzing causal linkages (Lin and Huynh, 2024). Discriminant and convergent validity were assessed prior to evaluating the study model hypotheses. Subsequently, the details were categorized by metrics representing the primary variables and their dimensions for the dependent variable (De La Rubia, 2019). The criteria for assessing closer suitability included factor loading exceeding 0.7; Average Extracted Variance (AVE) more than 0.4; and Cronbach's Alpha (CA) and Combined Reliability (CR) of internal consistency reliability surpassing 0.7 (De La Rubia, 2019).

The research objectives and the corresponding methodologies employed to attain them were detailed in Table 5. The initial objective was accomplished by Fuzzy Delphi Analysis, yielding 11 essential influencing factors belonging to 3 dimensions. The second objective was accomplished by SmartPLS analysis, yielding five acceptable hypotheses, as illustrated in Table 6.

Table 6 indicated that hypotheses H1, H2, H3, H5, and H6 were accepted, whereas hypothesis H4 was rejected. Nonetheless, the somatic situation was dismissed, which did not imply that it lacked an effect on the pupils' well-being. It suggested that somatic conditions remained a significant influencing factor for students' well-being, according to the opinions of experts. While Hypothesis 4 (predicting a significant relationship between somatic symptoms and egoism) was statistically rejected in our quantitative analysis, the persistent emphasis on somatic manifestations by clinical experts during the Delphi phase warrants methodological reflection. This divergence may stem from the following two interrelated factors: First, measurement granularity—experts contextualized somatic symptoms as embodied expressions of psychological distress (e.g., fatigue as intertwined with cognitive depletion), whereas our Likert-scale instrument may have captured these phenomena reductively, obscuring clinically meaningful nuances (Shen et al., 2022). Second, respondent awareness—students experiencing subsyndromal depression often lack somatic symptom attribution to mental health struggles, leading to underreporting in self-report measures (McGhie-Fraser et al., 2024). In comparison to the impact of somatic conditions, depressive symptoms, emotions, cognition, behavioral activity, and egoism were deemed more significant. These contributing elements must be given greater significance to foster students' mental resilience, mitigate depressive symptoms, and increase overall well-being.

Depressive syndrome impacts student well-being across emotional, cognitive, somatic, and behavioral dimensions (Park et al., 2023). College-based interventions show promise in addressing this relationship: resilience training fosters emotional growth (Cengiz et al., 2024), while positive psychology courses effectively teach happiness and resilience skills (Yang, 2023). A key protective factor, egoism - characterized by emotional regulation and curiosity - significantly enhances well-being by promoting autonomy, enjoyment, and stronger social connections (Leow et al., 2024). These findings highlight the importance of developing psychological resources to counter depressive symptoms while building student well-being (Park et al., 2023). Notably, while cognitive, emotional, and behavioral factors show limited direct impact on depression severity, they remain fundamental components of overall psychological well-being.

Depressive symptoms among Chinese college students, exacerbated by academic and socioeconomic pressures, necessitate enhanced mental health interventions. Institutions must prioritize psychological resilience and adaptive self-interest as key protective factors. By integrating these elements into curricula, campus policies, and extracurricular programs, universities can systematically strengthen students' coping capacities. The systematic implementation of complementary protective measures (resilience training and healthy ego development) would serve to both strengthen students' self-cognition and support enduring psychological well-being. This proactive approach addresses both immediate mental health needs and long-term adaptive functioning, aligning student support with China's evolving educational demands.

This research employed a mixed methodology via a survey for experts' feedbacks and a survey for students' data collection. The questionnaire design incorporated various characteristics from prior work while excluding factors such as emotional quotient, blunt capability, adversity quotient, and emotional support. In the future, additional factors should be considered to enhance the theoretical framework and construct a more robust architecture with clearly defined hierarchical dimensions. Owe to time limitation, the study employed the cross-section approach. Consequently, the transversal nature of the data collection approach obstructs the accurate evaluation of students' mental development. Consequently, the longitudinal technique is suitable as it can evaluate the robustness of the correlations.

In considering the study's limitations, future research should investigate associated variables to enhance the theoretical framework. A comparative analysis between two distinct sectors will elucidate the phenomena more comprehensively. Future research may adapt the proposed model in response to emerging scenarios to enhance its theoretical and practical implications. The forthcoming study should employ a longitudinal methodology to yield more comprehensive and dynamic data that cross-sectional studies have failed to provide, so enhancing the understanding of long-term trends and causal relationships.

Psychological resilience plays a crucial and beneficial role in relation to depressive symptoms and student well-being. To alleviate students' depressive symptoms and enhance students' well-being, the college should emphasize the development of psychological resilience through diverse interventions, including: (1) conducting cognitive-behavioral workshops to improve coping strategies, (2) incorporating mindfulness-based stress reduction (MBSR) programs into the curriculum, (3) creating peer support networks to enhance social connectedness, and (4) offering accessible counseling services featuring resilience-oriented therapy modules. The factor of egoism acts as a crucial protective moderator, significantly mitigating the adverse effects of depressive symptoms on students' psychological well-being. Thus, the college should prioritize the cultivation of students' egoism to protect against students' depressive feelings and obtain students' sense of well-being. Colleges can effectively equip students with self-preservation skills that protect against depressive tendencies by implementing structured programs (e.g., autonomy-supportive curricula, resilience-focused cognitive-behavioral training workshops, and ethical egoism discussions). This multifaceted approach enables students to manage academic pressures while preserving well-being, framing egoism not as narcissism but as a robust aspect of self-identity—a vital moderator in the link between depression and well-being.