Globally, labour markets are significantly transformed by globalisation, economic instability, technological advancement and shifting demographic trends (International Labour Organisation, 2023; World Economic Forum, 2023). In the process, countries are grappling across all income levels with persistent unemployment, escalating mismatches of skills and pressure to transition to knowledge-based economies (International Labour Organisation, 2023). In Africa, these global dynamics are compounded by challenges such as limited access to quality education and training, underemployment and slow industrial growth (African Development Bank, 2023; United Nations Economic Commission for Africa, 2023). In response, the African Development Bank’s Ten-Year Strategy (2024–2033) highlights the need to prioritise industrialisation, education and regional integration to stimulate inclusive economic growth and job creation across the continent (African Development Bank Group, 2024). Currently, rapid digitalisation and automation redefine the nature of work, placing increased pressure on both education systems and workforce development strategies to equip individuals with the skills and adaptability needed to thrive in a constantly evolving world of work (World Bank, 2024; World Economic Forum, 2023).

The World Economic Forum (2025) projects that 78 million new jobs could be created by 2030, which requires urgent upskilling efforts to match the changing demands of global economics. In South Africa, skills development, adaptability and inclusive employment strategies are particularly urgent because of unemployment, skills shortages and wage stagnation (Statistics South Africa, 2025; World Bank, 2025). In this context, the negative impact of rising inflation, fluctuating global trade markets, automation technologies and artificial intelligence (AI) cannot be ignored. According to the World Economic Forum (Di Battista et al., 2023), AI will impact over 60% of jobs globally within the next decade; thus, reskilling is important to adapt to the changing workplace. Furthermore, the high youth unemployment rate (45.5% in 2024) underscores a need for interventions that foster career adaptability, digital literacy and lifelong learning (Statistics South Africa, 2024).

However, while much of the existing literature focuses on enhancing adaptability resources such as skills and coping strategies, the outcomes of adaptation, like subjective career success, remain underexplored. Subjective career success is a critical adaptation result because it captures employees’ personal sense of fulfilment, alignment with values and perceived growth rather than relying solely on objective metrics such as income or job title (Heslin, 2005). Few empirical studies have examined how psychological dispositions (e.g. core self-evaluation) and adaptability resources jointly predict subjective career success in dynamic work environments. Addressing this gap is crucial for designing interventions that not only develop employees’ adaptability but also enhance their career satisfaction and engagement.

These challenges persist alongside a rapidly evolving higher education sector (International Labour Organisation, 2023; Universities South Africa, 2024), which demands digital and structural transformation, while shifting work models and global workforce trends shape career trajectories (eCampus News, 2025; Mlambo & Mpanza, 2024). In this context, employees must demonstrate career adaptation competencies and behaviours to navigate complex and evolving work environments. Instead of being reactive to external demands, career adaptation requires a dynamic process involving individual resources and contextual factors (Rudolph et al., 2017; Savickas, 2020), which positions adaptation as an ongoing process of aligning personal goals with environmental opportunities and constraints (Duarte et al., 2017; Lee et al., 2021). Therefore, we drew on Janeke’s (2023) refined version of the career construction model of adaptation (Rudolph et al., 2017), which includes personal dispositions, career adaptability resources, adaptive responses and subjective career success as outcomes of the adaptation process.

This study focuses on employees at a South African higher education institution because this sector is at the forefront of navigating digital transformation, funding constraints and evolving workforce expectations (Mlambo & Mpanza, 2024; Universities South Africa, 2024). Higher education institutions play a dual role in preparing graduates for the labour market while simultaneously managing the career development of their own diverse staff, who face similar challenges of job security, skills renewal and professional growth. Understanding how core self-evaluation and career adaptability influence subjective career success within this context provides actionable insights for organisational policies that foster staff retention, engagement and well-being. Therefore, this population offers a unique and relevant lens for examining the full adaptation process.

Previous research conducted in South Africa and other African contexts has emphasised the role of career adaptability resources, but a limited focus was placed on the outcomes of adaptation, particularly subjective career success. There is a growing need to explore the entire adaptation process, rather than isolating individual resources or traits (Diale & Van Zyl, 2021; Mashaba & Themane, 2022). Therefore, it is important to empirically test the relationship between core self-evaluation, career adaptability and subjective career success. This research was conducted within the framework of an adapted version of the career construction model of adaptation.

This study is grounded in Career Construction Theory (CCT), which views career development as a dynamic process shaped by how individuals construct meaning from their vocational experiences (Savickas, 2020). Central to this theory is the career construction model of adaptation, which outlines a process through which individuals adapt to changing career environments (Rudolph et al., 2017). Career adaptability is a fundamental construct in career construction theory, which suggests that individuals differ in their willingness and ability to develop career-related beliefs to engage in positive career-related behaviours that can lead to positive career responses and results (Savickas & Porfeli, 2012). Adaptation behaviours enable individuals to respond effectively to environmental changes, which promotes successful adaptation, experienced as career success (Rudolph et al., 2017; Savickas, 2020; Savickas & Porfeli, 2012).

The career construction model of adaptation (Figure 1) progresses from individual adaptive readiness (personal traits and dispositions), through the development of adaptability resources, leading to adapting responses (behaviours), and culminating in adaptation results such as career success or satisfaction (Rudolph et al., 2017).

Using this model as the theoretical framework allowed us to explore the interrelationships between core self-evaluation, career adaptability and subjective career success. In this model, core self-evaluation is positioned as a measure of adaptive readiness and serves as a key predictor of both career adaptability and subjective career success, while career adaptability is conceptualised as the individual’s adaptability resources, enabling flexible, forward-thinking responses to professional transitions and uncertainty. Thus, career adaptability functions as both an outcome of core self-evaluation and a predictor of subjective career success (Janeke, 2023). Subjective career success reflects the adaptation result, referring to an individual’s personal experience of fulfilment, satisfaction and meaning in their career progression. According to Janeke (2023), the adapted model assumes that employees’ capacity for core self-evaluation and career adaptability influence subjective perceptions of career success.

Subjective career success is considered to be an adaptation result that refers to an individual’s internal sense of achievement and satisfaction in their career (Heslin, 2005). Unlike objective indicators such as income or job title, subjective career success emphasises meaningful progression, personal fulfilment and alignment with individual values. For employees in academia and support services, subjective career success is often shaped by developmental opportunities, autonomy, work-life balance and recognition, influenced by personal agency and adaptability. The rapidly shifting labour market requires employees to demonstrate both career adaptability and core self-evaluation to achieve subjective career success, which reflects personal perceptions of career satisfaction, fulfilment and meaningful progression (Heslin, 2005). Given these dynamics, we examined whether core self-evaluation and career adaptability predict subjective career success as suggested by the career construction model of adaptation.

Career adaptability refers to an individual’s capacity to proactively manage a career by drawing upon adaptability resources and employing adapting responses to achieve positive adaptation results (Rudolph et al., 2017). It develops along four dimensions, namely concern, control, curiosity and confidence, which collectively empower an individual to plan for the future, take ownership of own development, explore opportunities and remain resilient in the face of uncertainty (Savickas, 2020). It is assumed that individuals with higher levels of career adaptability are more likely to engage in meaningful career planning, respond constructively to change and align personal goals with evolving labour market demands. In today’s rapidly shifting work landscape, those who develop stronger adaptability resources might be better positioned to sustain employability, experience greater job satisfaction and achieve long-term career success (Di Battista et al., 2023).

Core self-evaluation, a form of adaptive readiness, encompasses self-esteem, self-efficacy, emotional stability and locus of control, which shapes how challenges and opportunities in professional growth are perceived (Bono & Judge, 2003; Chang et al., 2012). Employees with positive core self-evaluations are more likely to engage proactively in career planning, persist in the face of setbacks and leverage development opportunities (Ng et al., 2021). Individuals with high core self-evaluation attain more complex jobs and adapt better to change, leading to higher job satisfaction, reduced stress and burnout and greater career success (Judge, 2009), while those with a negative core self-evaluation exhibit low life satisfaction and believe that their inabilities lead to failure and hinder the fulfilment of expectations (Özer et al., 2016).

Both core self-evaluation and career adaptability drive the ability to navigate changes; therefore, we assumed that both may have an effect on subjective career success, a measure of personal fulfilment in careers.

The model provided a meaningful lens to understand how personal dispositions, career adaptability and career-related outcomes interact within the career development process. It is also useful to offer practical recommendations for career development interventions in a South African higher education context, as improvement of core self-evaluation might foster subjective career success and retention.

We have employed a cross-sectional explanatory design from a positivistic paradigm. This design was chosen because it allows for the simultaneous estimation of multiple structural relationships using structural equation modelling (SEM), making it efficient for examining complex theoretical models in a single phase of data collection. Longitudinal or experimental designs, while valuable for establishing causal ordering, were not feasible because of resource and time constraints, as well as the practical challenges of tracking the same employees over multiple time points within a changing organisational environment. This type of design is limited by the extent to which it can prove causality; therefore, the causal ordering of variables was guided by two theories, namely career construction theory (Savickas, 2020) and the model of adaptation (Rudolph et al., 2017).

The population consisted of staff and contractors employed for 12 months or more at a South African higher-education institution. Because of the size of the population, we employed non-probability convenience sampling to select a group of participants (n = 242). Most of the participants were in middle adulthood (Table 1).

We conducted SEM, a statistical approach for analysing large samples, to test the proposed structural model and the hypothesised structural pathways. Accordingly, a power analysis (Wang & Rhemtulla, 2021) was conducted to confirm that the sample size was adequate to reliably reject the null hypotheses. As the requisite sample size in SEM varies depending on the implementation of item parcelling, we evaluated two distinct models, namely: (1) a holistic model wherein each latent variable was operationalised using three-item parcels and (2) a detailed model wherein the subdimensions of the subjective career success subscales were represented by individual items. Power analyses, configured with a power level of 0.8, a significance level of 0.05 and an anticipated effect size of 0.15, indicated a requisite sample size of n = 213. Therefore, our sample (n = 242) was deemed appropriate to reliably test the proposed hypotheses.

Three previously validated scales, namely the Subjective Career Success Inventory (SCSI), the Career Adapt-Abilities Scale (CAAS) and the Core Self-Evaluation Scale (CSES), were used to operationalise the variables under investigation. The scales were combined into a single survey and distributed to the target population using an online survey platform (EvaSys).

The SCSI measures eight dimensions underlying an employee’s perceived career success, namely authenticity, growth and development, influence, meaningful work, personal life, quality work, recognition and satisfaction (Shockley et al., 2016). Each factor is measured by three items presented with the stem ‘considering my career as a whole…’ and are measured on a five-point Likert scale ranging from ‘strongly disagree’ to ‘strongly agree’. The internal consistency for the scale (α = 0.83) was satisfactory, with support for the eight-factor structure of the model (comparative fit index [CFI] = 0.98; root mean square error of approximation [RMSEA] = 0.05) (Ibrahim & Amari, 2018). This scale was selected as the changing workplace necessitates a shift from objective to subjective factors when investigating employees’ views of their career success.

Career Adapt-Abilities Scale, developed by Savickas and Porfeli (2012), was used to measure career adaptability. The CAAS consists of four subscales, each with six items measuring concern, control, curiosity and confidence as psychosocial resources for managing career transitions and developmental tasks (Savickas & Porfeli, 2012). This scale was originally conceptualised as a multidimensional construct (RMSEA = 0.52; standardised root mean square residual [SRMR] = 0.40), but subsequent research has demonstrated strong support for a second-order (general) factor (Khampirat, 2024).

The Core Self-Evaluation Scale, developed by Judge et al. (2003), measures core self-evaluation. This 12-item scale consists of six positively worded items and six negatively worded items, which are rated using a five-point Likert scale. Positive wording examples include ‘I am confident I get the success I deserve’, while negative wording includes ‘I am filled with doubts about my competence’. This scale was selected as it demonstrated adequate internal consistency (α = 0.80) and strong construct validity (CFI = 0.93; RMSEA = 0.063; SRMR = 0.68) in various studies (Gu et al., 2015; Özer et al., 2016).

Ethical approval to conduct this study was obtained from the University of the Free State General/Human Research Ethics Committee (GHREC) of the institution, approval number: UFS-HSD2022/1620/22/23. The Human Resource (HR) Department played a pivotal role in upholding data protection standards during the research process. To solicit voluntary participation, the department used the official communication channel, namely email, to distribute invitations to employees. We did not have direct access to sensitive staff information, mitigating the risk of violating the provisions outlined in the POPI Act. Participation was voluntary, and the respondents could withdraw from the study at any time without any repercussions, as suggested by Sarantakos (2013). The participants were informed about the background and purpose of the study, and written informed consent was obtained before data were collected. No identifiable data were recorded to protect the identity of the participants.

The data were analysed in JAMOVI (Love et al., 2025), an open-source statistical analysis software package capable of advanced SEM. Cronbach’s alpha (α) was calculated to measure the internal consistency of the scales (Cronbach, 1951). Cronbach’s alpha assumes tau equivalence and unidimensionality, which are often violated during data analysis. Therefore, McDonald’s coefficient omega (ω) was also calculated because of less stringent assumptions surrounding factor structure and homogeneity (Enders, 2022). Both indices were used to determine whether the latent variables were successfully operationalised.

Convergent validity was assessed by computing the average variance extracted (AVE), which quantifies the mean proportion of variance explained (i.e. communality) across the items or item parcels of a given construct. As Hair et al. (2020) recommend, values exceeding 0.5 were acceptable. Discriminant validity was evaluated using the heterotrait-monotrait (HTMT) ratio of correlations. Although developed for variance-based SEM (Henseler et al., 2015), the HTMT metric has been adapted for covariance-based approaches (Afthanorhan et al., 2021). A threshold value of less than 0.85 is widely accepted as an indication of adequate discriminant validity (Hair et al., 2020).

Structural equation modelling, a second-generation multivariate analysis technique, was used to test the proposed model and the subsequent structural relationships between the variables (Hoyle, 2023). Confirmatory factor analysis (CFA) with robust maximum likelihood (RML) was used as the preferred estimation technique because multivariate normality assumptions were violated (Wulandari et al., 2021). Fit statistics were evaluated in relation to sample size and model complexity, as recommended by Kline (2023). The following indices were considered indicative of acceptable model fit: RMSEA, if < 0.08; CFI, if > 0.95; and Tucker–Lewis Index (TLI), if > 0.95. Less emphasis was placed on the SRMR, as it is often biased upwards for models with less than 12 observed variables (Hair et al., 2020).