Introduction

Rapid technological advancement, turbulence caused by global economic integration, and the effects of the recent pandemic have combined to create a work environment that is unprecedented in its complexity and uncertainty. First, rapid technological development and innovation have not only altered the character and structure of work but also created opportunities for those who are adaptable to these new technologies, while simultaneously increasing the competitive pressures on graduates (Benanav, 2020). Second, the interconnectedness of global economies implies that an economic setback in one nation could have repercussions in other nations, thereby increasing market volatility (D’Aguanno et al., 2021). Lastly, the detrimental global effects of coronavirus disease 2019 (COVID-19) have exacerbated an already difficult employment scenario, and the pandemic has significantly impacted daily lives and professions, casting alarming shadows over numerous career paths (Akkermans et al., 2020). Also, these effects have significantly influenced the labor market for Chinese graduates. As a result of the economic downturn, there has been a decrease in recruitment demand, an increase in employment supply, and an escalation in the challenges faced by Chinese graduates in securing employment (Mao et al., 2022).

For Chinese college graduates just entering the workforce, this sequence of challenges and changes presents an unprecedented dilemma. In this setting, perceived social support has emerged as a major factor in their career development. This social support, which may come from family, friends, or others, aids these graduates in enhancing their self-efficacy in career decision-making, that is, their confidence in their ability to execute career planning and decision-making duties successfully (Hou et al., 2019). Concurrently, this support may also aid in enhancing their career adaptability (CA), enabling them to navigate challenges and adjustments in their career trajectory more effectively (Öztemel & Yıldız-Akyol, 2021). This eventually contributes to an employee’s employment satisfaction (Takawira & Coetzee, 2019).

Therefore, it is essential to examine this topic in depth in order to gain a deeper understanding of the intricate relationships between these factors and to elucidate how graduates respond to these new demands and challenges. In light of this, the purpose of the present study is to investigate these questions in order to gain a deeper understanding of how, within this particular cultural and economic context, PSS interacts with career decision-making self-efficacy (CDMSE), career adaptability (CA), and job satisfaction, particularly for university graduates who are new to society.

Perceived social support

Social support includes factors closely related with an individual’s career development through his or her social connections, such as family members, relatives, colleagues, groups, and organizations (X. Wang et al., 1999). For individuals in the workplace, in addition to support from organizations, supervisors, and coworkers, it also includes relevant policies and measures. Even though college graduates in their early career stages have developed some career knowledge, they need more professional and social experience. They need to familiarize themselves with the professional world and life. Consequently, they will rely on the guidance of their social support system during the vocation selection, adaptation, and development processes. Two types of social support exist in the natural world. One is the support and assistance that exist in actuality and can be perceived through concrete and objective actions. Its existence manifests in both material and spiritual forms. The other is the individual’s belief that they will receive assistance and support for their situational problems, namely PSS, which has a connection with his prior subjective experience and emotion (Fu, 2015).

Social support is not only closely associated with an individual’s physical and mental health, but also with his or her career development (Fu, 2015). Wang (2014) found that social support possessed a substantial negative predictive role in career choice difficulties. According to Sun and Liu (2019), social support has a substantial beneficial predictive effect on CA. According to research of Han and Rojewski (2015), for college students who have worked for one year after graduation, their perception of school support positively affected their job satisfaction through adaptability. Moreover, through an investigation and analysis of special post teachers, it was concluded that social support could negatively predict attrition intention via job satisfaction (Li et al., 2021).

Career adaptability

The first to formulate the concept of CA based on job maturation and career maturity were Super and Knasel in 1981 (Song et al., 2023). Numerous academic researchers, including Pratzner, Goodman, Savickas, Ployhart and Bliese, and Johnston, have been interested in the concept of CA owing to its complex content and dimensions (Song et al., 2023). Currently, the definition proposed and refined by Savickas based on years of research is widely accepted: CA acts as a psychological asset that makes it possible for individuals to manage as well as activate their mental energy as they deal with career projects, difficulties, or shifts and when they experience career crises during career development (Savickas, 1997, 2005). Moreover, between 2008 and 2012, Savickas collaborated with experts from 18 countries and regions to develop operational definitions and measurement instruments for CA. Both qualitative and quantitative concepts were adopted to promote the operational process of CA (Savickas & Porfeli, 2012), and this international research demonstrates that experts and scholars from different countries accept and agree on the CA of the four-dimensional structure despite their different cultural backgrounds. These include concern, control, curiosity, and confidence.

Individual and situational factors influence CA (Rudolph et al., 2017), which is a form of social-psychological capital that accompanies individual career development to adapt to internal motivation and external environment changes (Savickas, 1997, 2005), and it is believed that there will be adaptive behaviors and outcomes (Zhou & Xie, 2022). Different demographic variables, such as age, gender, and level of education, influence the research outcomes of various academics. According to Rottinghaus et al. (2005) and Hirschi (2009), gender has no bearing on the development of CA. Jiang (2015) discovered that male students exhibited greater CA than female students. Using the Career Futures Inventory (CFI) in 2005, it was found that the level of CA among seniors was greater than the level of juniors, and this conclusion was confirmed (Rottinghaus et al., 2005). The research of Zhao and Xue (2010) and others demonstrates, however, that this was only sometimes the case. More studies have shown that personality traits such as the Big Five personality types (Rudolph et al., 2017), initiative, along with core self-evaluation, mental capability (Ohme & Zacher, 2015), individual emotion (Neureiter & Traut-Mattausch, 2017), and other could indicate the level of CA. Similarly, parenting style, social support, future time orientation (Öztemel & Yıldız-Akyol, 2021), and other factors may affect CA. Moreover, in studies on the effect of adaptive readiness in the psychological capital on life satisfaction (Pajic et al., 2018), the effect of parental autonomous support on academic engagement (Jiang et al., 2022), the effect of flexibility on workers’ life satisfaction (Topino et al., 2022), and others, adaptability in the workplace serves as a mediator.

Career decision-making self-efficacy

Taylor and Betz (1983) conceptualized career decision-making self-efficacy (CDMSE), which implements self-efficacy in making career choices, based on Bandura’s self-efficacy structure, cognitive behavior theory, and social learning theory. They believed that CSE is the decision-maker’s self-evaluation or belief in their capacity to fulfill their responsibilities as they make career decisions. It consists of the following five components: accurately evaluating one’s ability (including self-evaluation of one’s ability, professional interest, professional value, and self-concept), searching for professional information, correctly screening out jobs that match one’s characteristics with work characteristics, correctly planning the specific implementation of career decision-making, and having faith in one’s capacity to overcome obstacles. Over the years, scholars from various regions have conducted various studies on self-efficacy in career decision-making. Numerous studies have demonstrated a positive correlation between career-related familial support along with career selection self-efficacy (Lent et al., 2016).

In addition, recent research indicates that social support has a strong connection with CDMSE. Liu (2013) discovered a beneficial relationship between college students’ CDMSE and objective support, affective support, and utilization of support. Both objective and subjective social support can positively predict the CDMSE of college students. Other researchers have found similar results, with significant implications for individual career development. Social support and professional self-efficacy are significantly correlated in Chinese college students (Z. Wang & Fu, 2015); perceived support from educators had a positive association with self-confidence in making career choices and promotes individual adaptive career development (Di Fabio & Kenny, 2015). Moreover, the higher a person’s CDMSE, the greater his/her career maturity (Zhu & Miao, 2016).

Additionally, CDMSE has served as a mediator variable in a number of research relationships. Follow-up studies have shown that support from parents and teachers can predict job optimism one year later, in which the positive prediction is entirely mediated by CDMSE (Garcia et al., 2015). In addition, CDMSE acts as a mediator between career exploration and personality (Qu et al., 2015), as well as between familial support and adolescent career choice (Ginevra et al., 2015), among other variables.

Career satisfaction

The concept of career satisfaction (CS) is derived from the concept of job satisfaction, which cannot encompass CS. CS is a subjective indicator to measure career achievements, comprehensively reflecting employees’ satisfaction with their work achievements, salary, benefits, promotion opportunities, etc. It is closely related to employees’ guidance for career development, so it is more defined as the overall happiness people experience when choosing a career (Z. Wang & Long, 2009).

According to previous studies, primary factors influencing CS are concentrated on three levels: the individual, the vocation, and the organization. First, the perspective of the individual includes social population demographics, personality traits, and other factors. Demographic differences (Gattiker & Larwood, 1988; Pfeffer, 1991), gender, age, and marital status as well as individual personality characteristics (Frank et al., 1999) and positive psychological characteristics, including psychological capital and core self-evaluation, can have a substantial impact on the level of individual job satisfaction. Second, from a professional standpoint, working hours, tenure, work experience, work and career development pressure, and perception of excess qualifications negatively affect CS (Cheng et al., 2019).Third, from an organizational perspective, organizational support (Ng et al., 2005), employees’ perceived fairness in organizational career management, and organizations’ attention to employees’ career discovery (Herriot et al., 1994) can positively predict or influence employees’ CS. Rhoades and Eisenberger (2002) found that the support and attention of the organization can make employees more willing to work and take positive actions for their career development, thereby increasing their CS. Additionally, Hosseinkhanzadeh et al. (2013) discovered a significant correlation between organizational culture and satisfaction with work, and supportive organizational culture is essential for enhancing job satisfaction.

Research hypothesis

According to a summary of previous research, there is a strong connection across the four variables. Nonetheless, we can also find that previous researchers have paid insufficient attention to the relationship between these four factors, concentrating instead on the relationships between the two factors or one of the factors as an intermediary variable. Consequently, based on the analysis of previous related studies, this study is founded on Savickas’s career construction theory, and Rudolph et al.’s (2017) meta-analysis research results construct a chain intermediary hypothesis model between PSS, CA, CDMSE, and CS (Figure 1).

Figure 1
figure 1

The chained mediation of CA and career decision-making self-efficacy in the relationship between perceived social support and career satisfaction

Full size image

Hypothesis 1: PSS, CA, CDMSE, and CS of college graduates in China are significantly correlated with each other;

  • H1a: There is a significant association between PSS and CA of college graduates in China at the early stage of their careers.

  • H1b: There is a significant association between PSS and CDMSE of college graduates in China at the early stage of their careers.

  • H1c: There is a significant association between PSS and CS of college graduates in China at the early stage of their careers.

  • H1d: There is a significant association between CA and CDMSE of college graduates in China at the early stage of their careers.

  • H1e: There is a significant association between CA and CS of college graduates in China at the early stage of their careers.

  • H1f: There is a significant association between CDMSE and CS of college graduates in China at the early stage of their careers.

  • Hypothesis 2: The CA of early-career Chinese college graduates mediates the connection with PSS and CDMSE.

  • Hypothesis 3: The CA of early-career Chinese college graduates mediates the connection with PSS and CS.

  • Hypothesis 4: The CDMSE of early-career Chinese college graduates mediates the connection with PSS and CS.

  • Hypothesis 5: The CDMSE of early-career Chinese college graduates mediates the connection with CA and CS.

  • Hypothesis 6: There is a chain mediation effect between PSS, CA, CDMSE, and CS among Chinese college graduates at the early stage of their careers.

Methods

Participants

This study’s sample collection was authorized by the Scientific Research Office of Communication University of China, Nanjing. At the outset of the questionnaire, it is made abundantly clear that the survey is anonymous, that participation is voluntary, and that respondents may disengage and exit at any time. The data will be used exclusively for academic research and for no other purpose. After completing the questionnaire, there will be a random incentive drawing from 1 to 3 yuan.

Convenient sampling and snowball sampling were used to conduct online surveys. The sample is composed of college graduates from China who have worked for a minimum of 6 months, yet no more than 5 years after completing their initial studies, with no restrictions on the college type, major, or unit category. In November 2021, respondents participated in an online survey, and 700 questionnaires were distributed and recovered. Taking into account the consistency and completeness of the questionnaire responses and excluding outliers, 571 valid questionnaires (81.75%) were used for analysis. The sample consisted of 214 males (37.50%) employed in various regions of China, including 78 cities and various occupations. The participants’ essential information is displayed in Table 1.

Table 1 Summary of basic characteristics
Full size table

Measures

Perceive social support

The Perceived Social Support Scale (PSSS) devised by Zimet (1987) and translated by China scholar Jiang (X. Wang et al., 1999) was utilized to assess PSS. The scale emphasizes the self-understanding and sentiments of individuals regarding their social support, which is composed of three measurements: family, friend, and other support. There are 12 items in the scale, with 4 items in each measurement. Examples include “My family is willing to help me make decisions” and “I can discuss my problems with my friends.” In addition, it is a Likert 7 scale, ranging from 1 to 7, with higher scores indicating a higher level of support. Furthermore, this scale has obtained high levels of reliability and validity in Chinese research (Zeng & Huang, 2021).

Career adaptability

For the measurement of CA, the 12-item CAAS-SF scale revised by the team of Maggiori et al. was employed (Maggiori et al., 2017), and the expression was modified on the basis of the regional context. The gauge consists of four components: concern, control, curiosity, and confidence. There are three test questions for each dimension and a total of 12 elements. Examples include “I think about what my future will be like” (concern), “I count on myself” (control), “I look for opportunities to grow” (curiosity), and “I take care to do things well” (confidence). Adopting the Likert five-level assessment method, the greater the number, the greater the degree of conformity. Since its development in 2015, researchers in numerous regions have validated the reliability and validity of this scale (Song et al., 2023).

Career decision-making self-efficacy

The revised CDMSE-SF by Betz et al. (2005) was utilized to assess CDMSE. There were 25 questions covering five categories: self-evaluation, information-gathering, target-selecting, planning, and problem-solving. The scale employs a 5-point Likert scale, with higher scores indicating greater levels of CDMSE. Example queries include “Being able to identify what you value most in an occupation” and “You can find information about the school you graduated from.” This questionnaire’s reliability for homogeneity was 0.73. The reliability of retests was 0.80. The scale’s construct validity and reliability are high (H. Wang, 2021).

Career satisfaction

This study employed the Career Satisfaction Inventory (CSI) created by Greenhaus et al. (1990) and translated by Wang and Long (2009). The 5-item scale reflects satisfaction with career accomplishment, advancement, treatment, development, and abilities such as “I am satisfied with my progress in meeting my income target” and “I am satisfied with the success of my career.” On a Likert scale with 5 points, every item is scored; the higher the result, the greater the level of CS. The tool has received a great deal of backing for its reliability and validity and has been utilized by academics worldwide. This scale’s coefficient of internal coherence was greater than 0.80, indicating a high degree of dependability (Cheng et al., 2019).

Data analysis

Partial least squares (PLS) is a multivariate analysis method that incorporates the characteristics of principal component analysis and regression analysis. The primary purpose of PLS is to determine the relationship between two data sets. Particularly in structural equation modeling (SEM), PLS functions as an estimation method and is referred to as PLS-SEM; it is primarily employed for model estimation and forecasting. In this investigation, the model was constructed using PLS-SEM, and Smart-PLS version 3.3.7 was used to validate measurement and structural models. This is due to the fact that Smart-PLS is the most well-known statistical analysis instrument for PLS-SEM (Zhang, 2021).Moreover, as noted by Chin et al. (2003), PLS has less stringent sample condition requirements and does not require analysis data to adhere to a multivariate normal distribution, and investigation studies are generally not normally distributed.

Results

Typically, with PLS-SEM, both the measurement and structural models must be evaluated, and subsequent analysis can only be conducted if both models satisfy the conditions (Anderson & Gerbing, 1988).

Measurement model evaluation

When evaluating the measurement model in PLS-SEM, indicator reliability and internal consistency are typically evaluated using factor loadings, Cronbach α coefficient, and composite reliability (CR) values. In addition, convergent validity is determined by the value of average variance extracted (AVE), whereas discriminant validity is determined by the value of the heterotrait–monotrait ratio (HTMT) (Zhang, 2021). According to Zhang’s explanation of PLS-SEM, factor loadings must be more than 0.7 and significant at the 0.05 level, and Cronbach α and CR values of 0.7 or greater indicate higher internal consistency and reliability. When the AVE value of a construct exceeds 0.5, sufficient convergent validity is indicated. Moreover, examining the correlations between constructs and factor loadings allows for the evaluation of discriminant validity. It demonstrates discriminant validity between the constructs when the square root of the AVE for each latent variable (LV) is greater than the correlation coefficient between that LV and any other LV. In addition, an HTMT value less than 0.85 indicates excellent discriminant validity (Zhang, 2021).

Tables 2 and 3 display each item’s factor loading, reliability, and the convergence validity in this study’s measurement model, indicating that all items have adequate reliability, convergence, and differentiation effectiveness. In this study, PSS, CA, and CDMSE are second-order factors with three low-order dimensions (family, friend, and other), four low-order dimensions (concern, control, curiosity, and confidence), and five low-order dimensions (self-evaluation, information-gathering, target-selecting, planning, and problem-solving), respectively. As shown in Table 4, the second-order structure’s dependability and validity were also evaluated. The measurement of the second-order structure is also reliable and effective.

Table 2 Measurement model for the first-order constructs
Full size table
Table 3 Discriminant validity
Full size table
Table 4 Measurement model for the second-order constructs
Full size table

Structural model evaluation

After demonstrating the validity and dependability of the measurement model, the structural model can be evaluated to test hypotheses and address pertinent research questions and objectives. In PLS-SEM, the structural model is evaluated on the basis of various criteria, such as model fit, collinearity evaluation, path coefficients, coefficient of determinant (R2), effect size (f2), and predictive relevance (Q2).

Model fit

Model fit in Smart-PLS refers to the degree of fit between the model and observed data, i.e., how well the model corresponds with the actual data. Model fit indices evaluate the model’s quality by determining whether it adequately explains and predicts the observed data’s variability. Generally, square residual mean root (SRMR) is utilized for this evaluation. In general, an SRMR less than 0.08 or 0.1 indicates a decent fit of the model (Zhang, 2021). The SRMR value in this investigation is 0.059, meaning that it meets the criteria for fit.

Collinearity evaluation

Since the data were collected from a single source, the prevalence of multicollinearity among constructs can result in a common method bias issue. This can be determined by investigating multicollinearity in its entirety. In this technique, all variables are regressed on a common variable, and then the variance inflation factor (VIF) values are employed to determine the degree of multicollinearity. If VIF is less than 5, there is no significant multicollinearity between the constructs (Zhang, 2021). As shown in Table 5, the analysis revealed that the VIF is less than 5, indicating that single-source bias is not a significant concern for our data.

Table 5 The values of VIF, R2, f2, and Q2
Full size table

Path coefficients

It is essential to consider the standardized path coefficients (β-values) and their significance levels (t-statistics) when evaluating the structural model. The direct path loadings are regression coefficients with values ranging from −1 to 1 and represent the strength of the relationship between constructs. A value closer to +1 denotes a strong positive relationship, a value closer to −1 denotes a strong negative relationship, and a value closer to 0 denotes a feeble relationship, often not reaching statistical significance (Zhang, 2021). In addition, it is crucial to ensure that the path coefficients are statistically significant at the 0.05 level (Zhang, 2021).

Table 6 demonstrates that both t values (>1.96) and p values (0.05) satisfy the criteria, confirming that all hypotheses are accepted. Furthermore, the lower and upper limits of bias-corrected confidence intervals do not contain zero. Consequently, each of these direct associations is statistically significant. The variables can be ranked in terms of the strength and significance of their correlations as follows, in descending order: PSS has the strongest link with CA (β = 0.571), followed by CA with CDMSE (β = 0.477), CA with CS (β = 0.356), PSS with CDMSE (β = 0.273), CDMSE with CS (β = 0.201), and PSS with CS (β = 0.142). They are all positively correlated with one another. Therefore, support is provided for hypotheses H1a through H1f, establishing the groundwork for future investigations of hypotheses.

Table 6 Path coefficients and hypothesis testing direct effects
Full size table

Coefficient of determinant and effect size

In PLS-SEM, the coefficient of determination, also referred to as the multivariate correlation square (R2), and the effect size (f2) are common statistical indicators used to assess the model fit (Zhang, 2021).

R2 indicates the proportion of the variance in the dependent variable that can be explained by the independent variables and measures the model’s explanatory and predictive power. Chin (1998) established the threshold values of 0.19, 0.33, and 0.67 to categorize the effects as mild, moderate, and substantial, respectively (Zhang, 2021). As shown in Table 5, this study found that R2 for CA is 0.329 (weak), R2 for CDMSE is 0.455 (moderate), and R2 for CS is 0.367 (moderate). This indicates that PSS explains 32.9% of the variance in CA, 45.5% of the variance in CSE, and 36.7% of the variance in CS.

f2 is used to evaluate the extent to which exogenous variables influence particular endogenous variables. Typically, values of 0.02, 0.15, and 0.35 indicate mild, moderate, and strong effects of exogenous variables on endogenous variables, respectively (Zhang, 2021). In this model, the effect sizes of PSS, CA, and CDMSE on CS are 0.020, 0.104, and 0.035, indicating that their respective impacts are relatively weak (Table 5). PSS (f2 = 0.088) has a minimal effect on CDMSE, whereas CA (f2 = 0.289) has a moderate effect (Table 5). However, PSS (f2 = 0.490) has a substantial impact on CA.

Predictive relevance

In this study, the predictive relevance of the model was also assessed by calculating the cross-validated redundancy (Q2) values using blindfolding. Typically, when Q2 is greater than 0, it indicates that the model has effective predictive relevance for the provided latent variables, and a higher Q2 value suggests stronger predictive relevance (Zhang, 2021). As shown in Table 5, the Q2 values for the endogenous variables CA, CDMSE, and CS are 0.251, 0.353, and 0.282, respectively. These values all meet the criterion of being greater than 0, indicating that the model is predictive and effective.

Mediation effect

Exploring the mediating functions of CA and CDMSE in the relationship between PSS and CS is one of the objectives of this study. In the process of evaluating the mediation hypothesis, we followed the recommendations of Preacher and Hayes (2008) by using bootstrapping to test indirect effects. If the confidence intervals do not cross zero, we can conclude that there is a significant mediating effect. Table 7 shows that the indirect effects of each mediating relationship are significant, indicating the presence of mediating effects. Therefore, H2, H3, H4, H5, and H6 are supported. Consequently, the mediation analysis suggests that there is a significant chained mediation model between PSS and CS, with career adaptability and career decision-making self-efficacy mediating the positive relationship between PSS and CS.

Table 7 Hypothesis testing indirect effects
Full size table

Control variables and effects

In PLS-SEM, for controlling variables involving categorical data, multiple-group analysis is typically conducted for analysis and discussion (Zhang, 2021). In this study, the control variables include gender (female/male), household registration (rural/city), educational background (associate degree/bachelor’s degree), school category (public/private; undergraduate/junior college), major, years of work experience (less than 1 year/more than 1 year), and turnover experience (no turnover experience/turnover experience). These variables were all categorized as categorical data. In PLS-SEM, prior to analyzing the moderating effects of categorical variables on the model, it is necessary to evaluate the invariance of the between-group measurement model, which is a prerequisite for undertaking multiple-group analysis (Zhang, 2021).

By using the measurement invariance of composite models (MICOM) to assess the model, it was found that educational level (associate degree/bachelor’s degree), school type (public/private; undergraduate/junior college), major, and job turnover experience (no job turnover experience/job turnover experience) exhibited complete measurement invariance across the variables in terms of form, configuration, and variances. Gender (female/male), hometown (rural/city), and years of work experience (less than 1 year/more than 1 year) exhibited invariance in form and configuration, but they displayed variance differences. Specifically, gender had differential effects on CDMSE, while hometown and years of work experience had diverse impacts on PSS, with all other variables remaining invariant.

However, it is important to observe that full measurement invariance is a stringent and conservative requirement that is rarely completely met in empirical research. Some researchers consider that when at least two indicators for a construct have equal loadings and intercepts across groups, it demonstrates sufficient scalar invariance (Zhang, 2021). Overall, the control variables involved in this study did not substantially affect the between-group measurement model, indicating model invariance. Therefore, multiple-group analysis can be conducted.

Table 8 presents the results of multiple-group analysis for various control variables. Gender had a significant moderating effect on the path from PSS to CA. Educational background and school type had significant moderating effects on the path from CA to CS. Household registration had a significant moderating effect on the path from CA to CDMSE. Finally, years of work experience after graduation had a significant moderating influence on the path from CDMSE to CS. The other direct relationships were not affected by all control variables.

Table 8 Control variable effects
Full size table

Discussion

This study introduced CA and CDMSE to investigate the mechanism through which PSS influences CS. Based on the analysis of the collected data, a series of significant empirical results were obtained, confirming the hypotheses of this study and providing strong support for understanding the mechanisms underlying these relationships. The research findings indicate that PSS, CA, CDMSE, and CS have close and direct relationships. Furthermore, complex indirect relationships exist between them. Specifically, PSS not only has a direct impact on CS but also influences CS indirectly through CA and CDMSE. In addition, CA and CDMSE perform a mediating role in the relationship between PSS and CS, transmitting the influence of PSS to CS. Moreover, this study reveals that some control variables play moderating roles in these relationships.

Direct relationships between PSS, CA, CDMSE, and CS

Firstly, the study reveals that PSS has a positive effect on the CA, CDMSE, and CS of early-career Chinese university graduates (β = 0.571, p < 0.01 for CA; β = 0.273, p < 0.01 for CDMSE; β = 0.142, p < 0.01 for CS). This indicates that PSS functions as a significant factor influencing CA, CDMSE, and CS, aligning with previous research on the role of social support (Han & Rojewski, 2015; Li et al., 2021; Sun & Liu, 2019).

For Chinese university graduates preparing to enter or starting their professions, PSS exerts distinct pathways of influence. Individuals perceiving higher levels of social support tend to exhibit more proactive behaviors, such as actively seeking social resources, soliciting assistance, and effectively adapting to the challenges in their environment, thereby enhancing their CA and CDMSE. These factors interact and contribute to an increase in their CS. In contrast, individuals with inferior PSS may adopt more conservative strategies, exhibit less proactivity, and struggle to utilize available resources effectively. This situation may conceivably have adverse effects on their CA and CS.

Furthermore, the study also found that CA has a positive impact on CDMSE (β = 0.477, p < 0.01) and CS (β = 0.356, p < 0.01) among early-career Chinese college graduates. This discovery emphasizes the significance of CA in determining individuals’ career development. CA not only affects a person’s CDMSE and CS, but also their career trajectories by enhancing their CDMSE and CS. This result aligns with previous research findings (Lee & Jung, 2022; Rudolph et al., 2017).

Savickas emphasizes the importance of CA in career development and has proposed the theory of CA, which suggests that individuals cultivate CA in order to adapt to the incessant changes in their careers (Zhou & Xie, 2022). This theory underscores the significance of CA, and its various dimensions (concern, control, curiosity, and confidence) play a significant role in the early careers of Chinese college graduates. During this phase, they must address a number of career challenges, such as career selection, employment adaptation, and career advancement. Individuals with high CA are more likely to be actively concerned about their careers, possess greater control over their career trajectories, and exhibit a propensity to engage in learning and acquiring new knowledge and skills. They also approach career challenges with greater confidence. These behaviors and processes unquestionably enhance and improve their confidence and abilities in career decision-making, enabling them to better address career challenges and achieve their career objectives. Therefore, the positive impact of CA on CDMSE and CS is comprehensible.

Moreover, the study also revealed that CDMSE has a positive effect on CS (β = 0.201, p < 0.01), further emphasizing the significant function of CDMSE in influencing individuals’ CS. This discovery further emphasizes the significant function of CDMSE in influencing individuals’ CS. Graduates with high levels of CDMSE are more confident in their ability to make informed career decisions, plan their future career development, and solve career-related problems effectively. Thus, this has a positive influence on an individual’s CS, contributing to their overall career development.

Indirect relationships between PSS, CA, CDMSE, and CS

Using PLS-SEM modeling, this study found that, in the early careers of Chinese college graduates, CA positively mediates the effects of PSS on CDMSE and PSS on CS. Additionally, CDMSE also functions as a positive mediator in the relationship between PSS and CS, as well as between CA and CS. Moreover, there is an overall and effective chained mediation effect. The positive effect of PSS on CS is mediated by CA and CDMSE.

During the early stages of their careers, Chinese university graduates encounter numerous uncertainties and challenges, requiring them to adapt rapidly to the changing and complex professional environment. At this crucial moment, their PSS is crucial. Social support primarily comes from family, friends, and other significant interpersonal relationships, and is considered a valuable resource individuals need in their daily lives and work (X. Wang et al., 1999). PSS not only reduces graduates’ emotional stress by providing emotional support that enables them to better manage emotional fluctuations during the career adaptation and decision-making process but also provides essential information and advice regarding career adaptation and decision-making. This assists graduates in making wiser career decisions and taking appropriate action. This information and guidance increases their self-assurance and optimism in the face of career adaptation challenges, thereby enhancing their CS.

The various dimensions of CA play a significant role in the early careers of Chinese college graduates, as evidenced primarily by their career concern, control, curiosity, and confidence. Graduates with a high level of career concern are more motivated and self-aware in pursuing various forms of social support. They thoroughly comprehend the importance of this support in achieving their career objectives. This positive career perspective not only improves their ability to perceive social support but also increases their confidence in making career choices. Graduates with a strong sense of career control are more likely to actively pursue social support to meet their career requirements. They are better able to select appropriate career support resources and utilize them effectively to improve their CDMSE. Graduates with greater career curiosity are more inclined to interact with others and seek out new information and guidance. They exhibit a more positive attitude in perceiving social support and are more inclined to engage with family, friends, colleagues, and significant others to seek career advice and support. This active engagement strengthens their social support network and increases their CDMSE because they act with greater confidence. Graduates with higher career self-confidence are more likely to perceive the positive effects of social support. They interact with others with greater confidence and are more receptive to receiving support and counsel. This interaction strengthens their social support network and improves their CDMSE because they have the confidence to adjust to their career environment, set appropriate career objectives, and make intelligent decisions regarding their career development.

CDMSE also plays a significant role in the early careers of Chinese university graduates, predominantly evidenced in various aspects of their capabilities, including self-assessment, information gathering, goal setting, planning, and problem-solving. Typically, in the early phases of their careers, a high level of CDMSE makes university graduates more self-assured when evaluating their possession of the required skills and abilities. This confidence enables them to adapt proactively to new career environments, thereby enhancing their CA. This self-confidence is instrumental in confronting career challenges, enabling them to better surmount difficulties and thus increasing their CS. Furthermore, a high ability to gather information enables graduates to gain comprehensive insights into various aspects, such as the intended field, opportunities, support, and resources. This active information gathering and utilization are critical for perceiving various levels of social support effectively and increasing CA. It assists graduates in comprehending and adjusting to the career environment, thereby enhancing their CS. Graduates with strong goal-setting skills can intelligently set attainable career goals. Exceptional planning skills help them chart pathways to attain these objectives. Together, these competencies facilitate the development of CA, as well-defined objectives and effective plans aid graduates in adapting to new career environments. These skills also play an essential role in enhancing CS as graduates are more likely to realize their career aspirations. In addition, graduates with proficient problem-solving skills are better equipped to deal with obstacles during the CA process and find effective solutions. This proactive behavior helps strengthen their CA, as they have the confidence to surmount obstacles. Problem-solving skills also have a positive influence on career contentment because graduates can better appreciate success and a sense of accomplishment in their career journey. These positive behaviors and attitudes contribute to their career fulfillment.

CA is considered an adaptability resource, while CDMSE is seen as an adaptability response. These two factors are interconnected and are not only influenced by an individual’s adaptability preparedness but also have significant indirect impacts on adaptability outcomes (Zhou & Xie, 2022). This is especially pertinent for Chinese university graduates in the early phases of their careers, as they are entering the professional field and need to acclimatize to a new work environment. At this crucial juncture, their level of PSS directly impacts the extent and efficacy of their utilization of adaptability resources.

As discussed earlier, disparities in PSS capabilities will influence the actual support experiences, thus influencing their response behaviors. Chinese university grads in a state of survival have recently entered a new professional environment. This shift from a school learning environment, along with the distinctions in their positions, presents a transformation from the role of a pupil to that of a worker. If they cannot access and experience sufficient support from family, friends, peers, colleagues, and other external sources during this period, they may experience increased anxiety and isolation when confronted with career challenges. This emotional state can impact the individual’s utilization of adaptability resources, as well as diminish their enthusiasm for work, CA, and self-efficacy. They may begin to dispute their self-worth, ultimately resulting in dissatisfaction with their career status, which in turn impacts CS. Self-affirmation and self-satisfaction are more prevalent among individuals with a high level of social support. They tend to have a stronger sense of identification with the careers they pursue, which makes them more enthusiastic about their work, more satisfied with their careers, and ultimately aware of the significance and responsibility of their careers.

Control variables effect

The results of the multiple-group analysis in this study revealed the moderating effects of certain control variables on the relationships between variables. Specifically, in the early careers of Chinese college graduates, gender differences had a significant moderating effect on the relationship between PSS and CA. In this regard, the study found that the path coefficient difference between the female and male groups was −0.225 (p < 0.01), indicating that the impact of PSS on CS was substantially reduced for females compared with males. This result may reflect the intricate relationship between gender roles, societal expectations, and support systems among Chinese college graduates. It may also suggest the need to consider gender-specific support programs and intervention measures to better address the requirements of distinct gender groups.

In addition, the relationship between CA and CDMSE differed significantly among graduates from different geographic backgrounds. Specifically, the difference in path coefficients between rural and city groups was −0.208 (p < 0.05), indicating that CA had a substantially lesser effect on CDMSE in rural areas than in city areas. This distinction can be attributed to regional differences in the quality of primary education and the availability of resources. While rural development in China has been progressing significantly and the city–rural divide is progressively narrowing, the historical regional disparities can still have a lingering impact on individuals. This finding highlights the significance of contemplating regional differences in support and training programs in order to assist graduates from a variety of geographic regions more effectively.

In contrast, the number of years of post-graduation work experience significantly moderated the relationship between CDMSE and CS. The path coefficient difference between individuals with less than one year of work experience after graduation and those with over one year of work experience was −0.211 (p < 0.05). This suggests that individuals with less than one year of work experience after graduation had a substantially lesser impact of CDMSE on career satisfaction compared with those with more than one year of work experience. This finding may reflect that individuals’ perceptions and experiences of CDMSE endure significant changes as they accumulate work experience, particularly during their initial entry into the workforce. This discovery contributes to a greater understanding of the impact of the temporal dimension on early-career CS.

Finally, the control variables of educational background (associate degree versus bachelor’s degree) and school type (junior college versus undergraduate) had significant moderating effects on the relationship between CA and CS. In this regard, this study found that there were significant differences between individuals with an associate degree background and those with a bachelor’s degree background. The junior college group demonstrated a greater quantitative effect of CA on CS. This difference may be attributable to the distinct educational systems, resources, and faculty of various types of educational institutions, resulting in variations in the development of CA and CS among graduates from various backgrounds. This finding underscores the need to consider the requirements of pupils from diverse school backgrounds in educational policies and support.

Although, except for these influences, the control variables in this study did not have moderating effects on other relationships, the results of the multi-group analysis still underscore the significant role of control variables in the relationships among PSS, CA, CDMSE, and CS in the early careers of college graduates. These moderating effects provide valuable insights, aiding in the comprehension of differences between groups and providing support for individualized intervention and support measures.

Limitations and future study

This study, like all others, possesses limitations that must be considered. To begin with, it was a cross-sectional study, which calls for caution in interpreting causality. Although the data support the model’s hypothesized relationships, future research must confirm the precise directionality of these associations through longitudinal sampling. In addition, even though the type or field of work does not restrict the size of the sample of this investigation, the group size of certain types is insufficient; therefore, caution ought to be practiced while extending the results to the entire population. Future research should replicate these results with larger sample sizes. In addition, self-reporting methods were used to obtain the data for this study, which may have exposed them to reporting bias. Similarly, factors that may have influenced the participants’ subjective experience, such as family circumstances and marital status, were not investigated. To ensure that the results are pertinent to all occupational categories, future research may need to employ a sample that is more balanced and comprehensive.

Obviously, for the customization of support and intervention measures, training on various dimensions of PSS, CA, and CDMSE, as well as evaluating their efficacy, could be considered areas for further study.

Conclusion

In the swiftly changing, globalized world of today, the need to effectively manage constant change and uncertainty is becoming more and more pressing. In this context, this study reveals the positive and significant impact of PSS, CA, and CDMSE on the CS of Chinese college graduates in the early stages of their careers. Notable is the significant role that CA and CDMSE play as mediators in the relationship between PSS and CS. This means that our research findings emphasize the crucial role of CA and CDMSE in enhancing the positive relationship between PSS and CS. These findings have important practical implications for the development of CA support and intervention measures tailored to college students, aiming to meet the demands of the modern workplace.