Developing and validating a career competency framework for the Fourth Industrial Revolution: a modified Delphi study in South Africa

Abstract

Introduction:

The Fourth Industrial Revolution is reshaping work at pace, yet the career competencies required to navigate this transformation remain poorly understood, particularly in contexts marked by structural inequality and limited digital infrastructure. South Africa, with an unemployment rate of 32.7% and persistent skills mismatches, cannot rely on frameworks developed for conditions fundamentally different from its own. Critically, no validated, context-specific career competency framework currently exists for the South African formal professional sector engaged in 4IR-related career development an empirical, methodological, and contextual gap this study directly addresses.

Methods:

A two-round modified Delphi technique was employed with a purposively selected panel of South African experts drawn from industrial psychology, human resource management, organizational learning, and education across multiple provinces and sectors. Round 1 assessed definitional clarity of 38 candidate competencies derived from prior qualitative interviews, resulting in 23 retained, 12 refined, and five removed, with 35 competencies proceeding to Round 2. Round 2 applied dual thresholds for definitional precision and strategic importance across all 35 competencies.

Results:

Twenty-six competencies met all retention criteria and were organized into two tiers: 12 critical core competencies and 14 important contributors. The four highest-ranked competencies were career adaptability, ethical behavior, digital literacy, and continuous learning. Notably, three technology-related competencies, including artificial intelligence and digital agility, were excluded despite high importance ratings owing to insufficient definitional clarity, illustrating that perceived relevance cannot substitute for operational precision.

Discussion:

To the best of the authors' knowledge, this study presents the first expert-validated 4IR career competency framework developed specifically for the South African formal professional sector, addressing a significant empirical, methodological, and contextual gap in the career development literature. For curriculum designers and educators preparing graduates for AI-integrated workplaces, the framework offers both a validated set of teachable competencies and a concrete illustration of why AI-related skills resist easy formalization, a finding with direct implications for programme design and sequencing. The framework provides career psychologists with assessment-ready competency definitions, offers HR practitioners a sequenced developmental roadmap, and gives educators and policymakers an empirically grounded basis for curriculum alignment and skills policy reform, whilst demonstrating partial convergence with international workforce priorities.

Introduction

The Fourth Industrial Revolution (4IR) is fundamentally transforming the world of work globally, yet the career competencies needed to navigate this change remain poorly understood (Hirschi, 2018; Schwab, 2017). Beyond its disciplinary relevance to career psychology and human resource management, the study is deliberately framed within broader debates on South African labor-market inequality, digital exclusion, occupational skills mismatch, and the professionalization of career development practice. South Africa's professional labor market remains stratified by uneven access to digital infrastructure, historically rooted occupational segmentation, and a career development practice that is not yet fully professionalized as a regulated field; a validated competency framework therefore functions not only as a career-psychology instrument but as a potential lever for addressing structural exclusion from 4IR-aligned employment pathways (Ndung'u and Signé, 2020; Human Sciences Research Council, 2018). Whilst researchers have begun exploring this area (Chaka, 2020; Jo et al., 2024), the field is characterized by fragmentation, conflicting frameworks, and a significant lack of validated models that practitioners can rely on. This problem is especially acute in South Africa, where an unemployment rate of 32.7% (Statistics South Africa, 2026), ongoing skills mismatches, a persistent digital divide, and structural barriers including misaligned skills systems (Maree, 2009) and inadequate digital infrastructure (Wyrzykowski, 2023) shape career development in ways that most existing models do not address. The career psychology and human resource fields therefore require a prioritized, locally grounded framework built on structured expert consensus rather than imported assumptions. South African higher education institutions therefore face the compounded challenge of preparing graduates for an AI-integrated labor market whilst operating within a context that existing Global North frameworks systematically misrepresent. The career psychology, human resource, and education fields consequently require a prioritized, locally grounded framework built on structured expert consensus rather than imported assumptions.

The origins of existing frameworks compound this problem. Most career competency frameworks were developed in the United States (Akkermans et al., 2013; Kuijpers and Scheerens, 2006), the United Kingdom (Francis-Smythe et al., 2012), and Australia (Sargent and Domberger, 2007), rendering their relevance to South Africa's unique cultural, economic, and technological context debated (Naidoo et al., 2024; Savickas, 2005). Three distinct gaps therefore exist in the literature: (1) an empirical gap as there are no validated career competency framework exists for the South African formal professional sector (Barnes et al., 2022; Coetzee and Schreuder, 2020); (2) a methodological gap as existing frameworks have not undergone Delphi-based expert validation or structured prioritization that would enhance their practical utility (Bal, 2024; De Vos et al., 2020); and (3) a contextual gap as frameworks originating in the Global North embed assumptions about digital access, labor market stability, and institutional trust that do not reflect South African realities (Gartoumi and Tekouabou, 2024; Laher and Cockcroft, 2018; Sultana, 2024). For curriculum designers, this third gap is particularly significant: a framework that misrepresents the context in which graduates will work cannot reliably guide programme design.

This study directly addresses these gaps through a two-round modified Delphi study (Avella, 2016; Hasson et al., 2025) involving career psychology, human resource management, and 4IR experts across multiple South African provinces and sectors. The outcome is the first expert-validated, context-specific 4IR career competency framework for South Africa, designed for use by career psychologists, HR professionals, educators, and policymakers.

The study pursued three objectives: first, to refine an initial list of career competencies and establish expert consensus on their relevance through a structured Delphi process; second, to assess and prioritize the relative importance of validated competencies for long-term career sustainability within the South African 4IR context; and third, to develop a finalized, expert-validated framework of 4IR career competencies organized into coherent domains.

The scope of the study is deliberately confined to South Africa's formal professional sector. The Delphi panel, and the resulting framework, do not extend to mining, agriculture, retail, construction, public health, or the informal economy, sectors that collectively account for the majority of South African employment. This restriction reflects the purposive sampling criteria used to recruit the expert panel and is revisited in the Limitations section, together with its implications for the framework's generalisability.

Theoretical background and literature review

Developing a credible career competency framework is a methodologically complex task. Campion et al. (2011) describe competency modeling as both a science and a craft requiring systematic data collection, expert input and rigorous validation. Despite increasing scholarly interest in career competences (Akkermans and Kubasch, 2017; Jo et al., 2024), a recurring issue remains. Frameworks are often proposed without sufficient validation before practitioners are exposed to using them.

The theoretical origins of competency as a construct trace back to White (1959), who described it as the ability to interact effectively with one's environment. McClelland (1973) later argued that behavioral traits, rather than intelligence alone, predict real-world career success. Boyatzis (1982) expanded this view to include emotional intelligence and interpersonal skills as essential to career success. These contributions collectively established that the competency construct is multidimensional, context sensitive and developable. These are the qualities that directly influence how frameworks should be created and tested.

A useful distinction has emerged in the literature between competency identification and validation (Batt et al., 2021; Megahed, 2018). Identification involves generating candidate competences through literature review or empirical inquiry. Validation goes further. It tests whether those competences are genuinely relevant, important and applicable across a target population. Without this step, frameworks risk reflecting researchers' assumptions rather than practitioners' lived experiences (Campion et al., 2011; Spencer and Spencer, 1993).

This distinction is especially significant in South Africa. Both the Human Sciences Research Council (2018) and the Presidential Commission on the Fourth Industrial Revolution (2020) have called for evidence-based frameworks that reflect local conditions. Yet no validated framework currently exists for the South African formal professional sector (Barnes et al., 2022; Coetzee and Schreuder, 2020). This study directly addresses that gap.

The intelligent career framework

The study adopted the Intelligent Career Framework (ICF; Arthur et al., 1995; DeFillippi and Arthur, 1994) as a contextualizing conceptual framework rather than a direct design scaffold. The ICF was developed for modern careers characterized by mobility and individual agency. Its six dimensions—knowing why (values and motivation), knowing how (skills and expertise), knowing whom (networks and relationships), knowing what (opportunities and challenges), knowing when (timing and transitions), and knowing where (environmental positioning)—provided a broad conceptual map for understanding the multidimensional nature of career competencies. However, the Phase 1 interviews and Delphi rounds were not structured around the ICF's six dimensions; rather, the framework informed the interpretive lens applied to the validated competencies in the Discussion. Several features make this framework well-suited for the present study. Its six dimensions offer a comprehensive overview of career competency areas without being tied to specific occupations, making it suitable for identifying transferable competencies across diverse careers (Arthur et al., 1995). It considers both individual and contextual factors, which is essential when examining careers shaped by South Africa's complex socio-economic environment (Arthur et al., 1995). Its established use in international research also supports cross-contextual comparisons of the validated framework (Beigi et al., 2018; Crowley-Henry et al., 2023).

Its limitation is that it predates the Fourth Industrial Revolution era and does not fully encompass competencies such as digital agility, human-computer collaboration, or digital literacy (Bouwmans et al., 2024). Therefore, the framework was utilized as a conceptual model rather than a definitive one, with expert validation rounds designed to identify career competencies beyond its original dimensions where necessary (Beigi et al., 2018; Crowley-Henry et al., 2023).

Competency prioritization and competency sustainability

Identifying relevant competencies is valuable, but practitioners also need to understand which competencies matter most, especially when working with populations facing resource constraints. Career counselors, HR professionals, and educators in South Africa require clear developmental priorities, not just a comprehensive list (Spurk et al., 2019). Prioritization is therefore as important as validation in creating a framework that is practically useful.

Career Sustainability Theory offers a valuable perspective on how professional development occurs across different stages (De Vos et al., 2020). According to De Vos et al. (2020), the construct suggests that sustainable careers are those that are meaningful, healthy, and financially viable over time and in various contexts. This view indicates that different competencies are valued differently at various career phases. Foundational skills are essential at the initial entry into the labor market. These are not the same as the advanced skills required to maintain, advance, or transition in a career, especially amid technological disruption (Bal, 2024; McMahon and Abkhezr, 2025).

The Fourth Industrial revolution adds an extra layer of complexity. The World Economic Forum (2025) predicts significant skills disruptions across various sectors, driven by both automation and augmentation (Lei and Kim, 2025; Willcocks, 2024). In South Africa, these global challenges intersect with local structural issues, including high unemployment, unequal access to digital infrastructure, and ongoing educational inequalities (Gonese and Ngepah, 2024; National Planning Commission, 2012), which influence the agency and feasibility of specific competency investments. Notably, research indicates that some capabilities persist through technological changes. Deloitte (2019) highlights adaptability, learning agility, and interpersonal effectiveness as enduring human competencies that maintain value even when technical requirements evolve. Neamtu et al. (2025) and Zenk et al. (2024) similarly stress the significance of meta-competencies, the higher-order competencies that enable workers to acquire, deploy, and adopt other skills as demand shifts (Munsamy et al., 2023). Therefore, the prioritization aspect of the study was designed to address both immediate developmental needs and the long-term foundations for career sustainability.

A complementary theoretical perspective is provided by Career Construction Theory (Savickas, 2002, 2013, 2021), which emphasizes adaptive meaning-making and narrative identity construction as central to career development. Whereas the ICF foregrounds the accumulation and deployment of career capital across six knowing dimensions, Career Construction Theory focuses on how individuals construct subjective career meaning through adaptation to changing vocational contexts. The two frameworks rest on different epistemological assumptions: the ICF adopts a broadly rational-agentic view of career competence, whilst Career Construction Theory emphasizes interpretive and constructionist processes. Both perspectives inform the interpretive approach adopted, where the validated competencies are examined through both lenses. The inclusion of Career Construction Theory is particularly appropriate given this study's emphasis on career adaptability, which Savickas (2013) identifies as a core meta-competency enabling individuals to manage vocational transitions and construct coherent career narratives amid disruption.

The South African context

Career competency frameworks must be embedded within the context in which they are applied (Einarsdóttir et al., 2010; Savickas, 2005). Three contextual factors were especially influential in shaping the development and prioritization of career competences in this study. First, South Africa's skills development system remains primarily focused on job-specific training rather than transferable competency-based development (Department Higher Education and Training, 2014; Hattingh, 2016). This results in a structural mismatch between educational outputs and adaptable competency profiles that 4IR employers seek (Samuel and Moagi, 2022; Seti et al., 2025). A validated framework based on expert consensus could guide reform of this system, but only if it considers the institutional constraints that influence how the South African workforce development functions.

Second, even the digital infrastructure continues to restrict career competency development for large parts of the workforce. Many South African workers lack reliable Internet access, limiting their ability to acquire digital skills, participate in ongoing learning or access digital labor markets (Ndung'u and Signé, 2020; Sikhakhane and Lubbe, 2005). The Human Sciences Research Council (2018) has emphasized that the 4IR transformations in the South African context require concurrent attention to infrastructure development and skills enhancement. Competency frameworks that assume universal digital access do not reflect this reality.

Third, the increasing use of AI-driven recruitment tools and digital hiring platforms introduces new forms of labor market exclusion. These systems tend to favor candidates with strong digital literacy and access (George et al., 2025; Marchal et al., 2007), risking the systemic exclusion of those already disadvantaged. A prioritized career competency framework for South Africa must therefore address skills that enable labor market participation, including digital navigation and platform literacy, alongside those related to job performance and career progression.

The three aforementioned factors were not treated merely as background context but were integrated directly into the expert panel composition and the rating criteria used throughout Delphi rounds. This approach, in short, ensures that the consensus outcomes of the study reflect the practical realities of the South African context rather than being assumptions imported from elsewhere.

Taken together, four theoretical and contextual strands inform this study's interpretive approach and the Intelligent Career Framework provides the structural account of multidimensional career capital; Career Construction Theory explains adaptive meaning-making and narrative agency during career transitions; Career Sustainability Theory accounts for why validated competencies must support long-term, healthy, and financially viable careers rather than short-term task performance; and the 4IR and South African contextual literature establishes the technological and labor-market conditions against which competency relevance and feasibility were judged. These strands are integrated in the discussion to interpret the validated framework, rather than having directly structured the Delphi instrument itself.

Materials and methods

Delphi consensus-building process

Rationale for Delphi methodology

The modified Delphi technique was employed to achieve expert consensus on 4IR career competencies. Originally developed by Dalkey and Helmer (1962), the Delphi method facilitates structured group communication for solving complex problems without direct interaction, thereby eliminating biases inherent in group dynamics (Hasson et al., 2025; Linstone et al., 2002). This method is particularly effective for establishing consensus in domains characterized by limited empirical evidence and where expert judgement is essential (Niederberger and Köberich, 2021; Schifano and Niederberger, 2025). The Delphi technique's capacity to gather geographically dispersed expert opinions whilst preserving anonymity makes it especially suitable for emerging phenomena such as 4IR competencies (Nasa et al., 2021; Skulmoski et al., 2007).

Expert panel composition

Consistent with the formal-professional-sector scope stated in the Introduction, purposive sampling was employed to recruit experts meeting stringent inclusion criteria: (a) minimum master's degree in industrial psychology, psychometrics, human resource management, or related disciplines; (b) minimum 5 years' professional experience with career competencies during the 4IR period (2016 to present); (c) active employment in advisory, technical, or leadership roles involving career development, competency assessment, or 4IR implementation; and (d) professional registration with relevant bodies such as the Health Professions Council of South Africa (Blasco-Blasco et al., 2025; Khodyakov et al., 2023; Okoli and Pawlowski, 2004).

Panel size was determined following methodological guidance for homogeneous expert groups, which recommends 10–15 participants (Delbecq et al., 1975), with empirical research supporting 8–23 participants for specialized professional domains (Khodyakov et al., 2023). Accounting for anticipated attrition rates of 20%−30% (McMillan et al., 2016), 30 experts were initially invited. Seventeen experts participated in Round 1 (57% response rate), with 13 completing Round 2 (76.5% retention rate). This panel size aligns with established guidelines and ensures adequate representation whilst maintaining engagement quality (Belton et al., 2021; Nasa et al., 2021).

The demographic composition of the expert panel is presented in Table 1. The panel comprised 17 members, of whom 13 were women (76.5%) and four were men (23.5%). The majority were in their 40s, with nine experts (52.9%) aged between 40 and 49. Three experts (17.6%) were in their 30s and four (23.5%) were in their 50s, indicating that most panel members were experienced mid-career professionals.

Of the 17 experts who participated in Round 1, 13 (76.5%) continued to Round 2, whilst four (23.5%) did not. This retention rate falls within acceptable ranges for Delphi studies, where some attrition between rounds is expected owing to panelists' time constraints and competing commitments (McMillan et al., 2016). The 38 candidate competencies presented to the Delphi panel in Round 1 were derived from Phase 1 of the broader study (Mtshali and Dhanpat, 2026) which comprised semi-structured qualitative interviews with 15 South African professionals purposively selected for their direct experience of 4IR-related career transitions. Participants were drawn from industrial psychology, human resource management, education, and technology-intensive sectors across three South African provinces (Gauteng, Western Cape, and KwaZulu-Natal). Interviews explored participants' experiences of competency demands in 4IR work environments, the career capabilities they regarded as essential for navigating technological disruption, and the contextual factors shaping competency development in South Africa. Interview transcripts were analyzed using thematic analysis (Braun and Clarke, 2006), yielding 38 candidate career competencies with preliminary definitions. These competencies served as the input for the Delphi consensus process described below.

Delphi round 1: competency clarity assessment

The Round 1 online questionnaire, distributed via SurveyMonkey, presented 38 career competencies derived from qualitative interviews conducted in Phase 1 of the broader study. Experts rated each competency descriptor on a 5-point Likert scale (1 = not at all clear, 5 = extremely clear) to assess definitional clarity (Joshi et al., 2015). Mandatory comment fields enabled experts to propose modifications, clarifications, or additions to competency descriptors when ratings indicated insufficient clarity (scores of 3 or below). This modified Delphi approach integrated quantitative ratings with qualitative feedback, facilitating collaborative refinement of the competency framework (Bauls and Turoff, 2011; Linstone et al., 2002). Participants were afforded 4 weeks to complete the questionnaire, balancing thoughtful consideration with momentum maintenance (Nasa et al., 2021).

Comprehensive descriptive statistics were computed for each competency: mean scores, standard deviations, medians, interquartile ranges (IQR), and percentage agreement (proportion of responses rated 4 or 5). Consensus thresholds, established a priori following systematic reviews of Delphi methodology (Diamond et al., 2014; Nasa et al., 2021), classified competencies as: (a) retained, meeting all criteria of 70% or greater agreement, IQR of 1 or less, mean of 3.50 or greater, and median of 4.0 or greater; (b) refined, meeting partial criteria of 50%−69% agreement, or mean of 3.50–4.0 with moderate dispersion (IQR = 1); or (c) removed, failing to meet minimum standards of less than 50% agreement, mean below 3.50, IQR of 2 or greater, or median below 3.0 (Terbraak et al., 2024; von der Gracht, 2012).

Expert comments underwent systematic thematic analysis using directed content analysis to identify specific clarity issues, proposed modifications, and rationales for ratings (Brady, 2015; Sekayi and Kennedy, 2017). Quantitative and qualitative analyses were integrated to inform structured decision-making: competencies meeting retention thresholds were carried forward unchanged; those meeting refinement criteria underwent systematic revision based on expert feedback; and those failing minimum standards were eliminated from the framework (Beiderbeck et al., 2021).

Delphi round 2: validation and importance ranking

Round 2 employed a three-phase assessment design encompassing: (a) clarity validation of refined competencies (5-point scale: 1 = not at all clear, 5 = extremely clear); (b) ambiguity assessment of retained competencies, using the same 5-point scale, where higher scores indicated greater ambiguity and were reverse-scored for analysis so that a mean of 4.0 or greater indicated low ambiguity (that is, high clarity); and (c) importance ranking of all competencies (7-point scale: 1 = not important at all, 7 = extremely important). This comprehensive approach enabled simultaneous validation of definitional precision and establishment of strategic importance hierarchies (Rajhans et al., 2021; Yao et al., 2025).

Dual threshold criteria, established a priori (Diamond et al., 2014; Holey et al., 2007; Zhang et al., 2025), required competencies to demonstrate consensus across two dimensions. For definitional consensus, refined competencies required: mean of 4.0 or greater, coefficient of variation (CV) of 30% or less, percentage agreement of 85% or greater (responses rated 4 or 5), and IQR of 1 or less. Retained competencies required: mean of 4.0 or greater, CV of 30% or less, and IQR of 2 or less. For strategic importance, all competencies required: mean of 5.5 or greater on the 7-point scale, CV of 40% or less, frequency agreement of 85% or greater (responses rated 6 or 7), and IQR of 2 or less. The specific threshold values were informed by the ranges reported in systematic reviews of Delphi methodology (Diamond et al., 2014; Nasa et al., 2021) and adapted for the present study. Diamond et al. (2014) report that percentage agreement thresholds in published Delphi studies commonly range from 70 to 97%, while Nasa et al. (2021) recommend that consensus criteria be established a priori and calibrated to the study context rather than applied mechanically. The definitional clarity threshold of 85% agreement for refined competencies was set at the upper end of this range to reflect the stringent operationalisation demands of a career competency framework intended for assessment and development use. The strategic importance frequency agreement threshold of 85% was set to ensure that retained competencies had near-unanimous endorsement of their essentiality. These adapted thresholds are consistent with recent Delphi studies employing dual-criterion validation in health and social science contexts (Zhang et al., 2025; Yao et al., 2025).

For clarity and ambiguity assessments, descriptive statistics (means, standard deviations, variance, mode, median, IQR, CV) were computed. For importance rankings, comprehensive statistics enabled hierarchical prioritization of competencies for 4IR career development initiatives. CV served as a standardized consensus measure, with lower values indicating stronger expert agreement (Holey et al., 2007).

Data collection and quality assurance

Both Delphi rounds employed SurveyMonkey for electronic distribution, preserving the anonymity essential to Delphi methodology by preventing participant awareness of fellow panelists' identities or responses (Niederberger et al., 2024). Automated reminders facilitated response tracking. Data quality protocols encompassed completeness assessments, response pattern evaluations, and rating scale verification. Personalized follow-up communications maintained engagement whilst acknowledging expert contributions (Belton et al., 2019; Khodyakov et al., 2023).

Expert panel validation employed stringent verification procedures, including credential confirmation, experience validation, and professional registration verification. The 4IR digital connectivity context enabled triangulation of self-reported expertise through institutional websites, LinkedIn profiles, and academic publication databases. Geographic and sectoral diversity monitoring ensured comprehensive representation throughout both rounds (Hasson et al., 2025).

Methodological rigor and limitations

Several strategies mitigated potential limitations. Panel heterogeneity across sectors, roles, and contexts reduced perspective bias whilst maintaining homogeneity in professional expertise (Bonaccorsi et al., 2020; Mukherjee et al., 2015). Individual questionnaire distribution and anonymity countered groupthink, halo effects, and social pressure, enabling independent expert judgement (Khodyakov et al., 2023; Nasa et al., 2021). Transparent inclusion criteria, systematic recruitment procedures, and comprehensive demographic documentation enhanced expert selection rigor (Hasson and Keeney, 2011).

The researcher's professional familiarity with some participants, whilst potentially introducing bias, conferred methodological advantages including enhanced response rates and contribution depth. This approach aligns with purposive sampling recommendations for Delphi studies, wherein researcher knowledge of the expert community facilitates identification of qualified participants (Hasson and Keeney, 2011). Supervisor examination of qualitative analyses and systematic response monitoring throughout both rounds ensured analytical reliability and participant engagement quality (Niederberger and Spranger, 2020).

The 13-expert panel size in Round 2, whilst within recommended ranges for homogeneous expert groups and demonstrating strong retention (76.5%), represents a limitation regarding generalization. However, the panel's geographic, sectoral, and role diversity, combined with stringent consensus thresholds, strengthens confidence in the framework's applicability across South African 4IR contexts.

The decision to conclude the Delphi process after two rounds warrants methodological justification, particularly given that 42% of competencies were modified, removed, or reconceptualised following Round 1. Several considerations supported this decision. First, the modified Delphi design employed in this study differs from the traditional iterative Delphi approach in which successive rounds aim for incremental opinion convergence. In the present study, Round 1 served a diagnostic function, identifying competencies requiring definitional revision, while Round 2 applied the most demanding dual-threshold criteria simultaneously to all 35 competencies, including both refined and newly created constructs. Stability was therefore assessed not merely in terms of opinion consistency but through a stringent dual-criterion test requiring both definitional precision and strategic importance. Second, the Round 2 statistical indicators demonstrated stability: the mean coefficient of variation across retained competencies was 19.4% (well below the 30% threshold), and 26 of 31 eligible competencies met all retention criteria. Third, the methodological literature on modified Delphi studies recognizes that the stopping rule should be determined by the research design rather than by a fixed number of rounds (Hasson et al., 2025; Niederberger and Spranger, 2020). Where a modified design employs pre-established quantitative thresholds as the primary stability criterion, as in this study, reaching those thresholds constitutes a defensible basis for concluding the process.

Results

A modified two-round Delphi technique was employed to develop and validate a career competency framework for the Fourth Industrial Revolution (4IR). Phase 1 qualitative interviews generated 38 competencies, which were subsequently subjected to expert consensus evaluation. Retention criteria were established a priori in line with established Delphi methodology (Diamond et al., 2014; Nasa et al., 2021); full threshold specifications are detailed in the Methods section. The process yielded a final validated framework of 26 competencies.

Round 1: initial consensus evaluation

Three key patterns emerged from Round 1 (see Table 2). First, interpersonal and cognitive competencies achieved the strongest consensus: collaboration and artificial intelligence both reached 100% agreement, whilst communication skills, critical thinking, creativity, professional social media, and cognitive load management all achieved 94.1%. Second, technology-related competencies showed moderate consensus, suggesting familiarity with the concepts but uncertainty about their operationalisation at the individual competency level. Third, self-referential constructs including self-evolution, emotional intelligence, and diverse knowledge showed the weakest consensus (agreement 52.9%−56.3%), reflecting definitional disagreement rather than rejection of their underlying importance. Overall, 23 competencies (60.5%) met retention criteria, 12 (31.6%) required refinement, and five (13.2%) were recommended for removal.

Expert consensus was categorized as strong (90% or greater), moderate (75%−89%), weak (65%−74%), or absent (below 65%), consistent with established thresholds (Diamond et al., 2014; Nasa et al., 2021). Eight competencies (21.1%) achieved strong consensus, 15 (39.5%) demonstrated moderate consensus, eight (21.1%) achieved weak consensus, and four (10.5%) fell at or below the consensus threshold.

Qualitative refinement following round 1

Expert qualitative feedback informed targeted refinements to 12 competencies and led to the creation of two new constructs. All qualitative responses were analyzed by the primary researcher using directed content analysis; the supervisor independently reviewed all coding decisions and refinement rationales to ensure analytical consistency. Where the primary researcher's and supervisor's interpretations of an expert comment differed, these were resolved through discussion until consensus on the coding decision was reached.

Proactivity was reconceptualised as career proactivity after several panelists distinguished between self-directed career development and workplace efficiency. Expert 5 stated: “I think there should be differentiation between being proactive for self (learning new skills and career growth) and proactive at work (improving processes/efficiencies),” whilst Expert 7 recommended that “career proactivity should be preferred” to avoid conflating a career competency with a general personality trait. The latter component was operationalised as a discrete new competency, process improvement skills.

Technological adaptability was re-labeled human-technology collaboration after experts identified overlap with existing competencies. Expert 9 asked: “How is this different to adaptability, and continuous learning?” and Expert 16 noted it was “not clear what we are measuring here,” prompting a refined focus on integrating automated solutions with human judgement.

Digital literacy was refined following Expert 3's recommendation that the definition focus on “the ability to use technology to find, evaluate, create and communicate information effectively.” This refinement also prompted the creation of digital agility as a separate construct after Expert 11 asked: “Where does digital agility come in?” Broadening skill sets was re-labeled cross-functional skills, with the description narrowed to focus on collaboration across functional boundaries, after Expert 16 observed that the original definition was “messy and all over the place” and needed clarity “to allow assessment of a distinct competency.”

Boundary-setting and work-life integration were refined after Expert 13 observed: “There seems to be conceptual overlap between work-life integration and boundary-setting.” Contextual business intelligence was consolidated into business acumen after Expert 5 noted it was “a combination of business acumen and market intelligence” and Expert 9 questioned: “Not sure how this is different to the previous competency or business acumen.” The descriptions of resilience, continuous learning, and business acumen were similarly refined to reduce conceptual overlap and improve definitional precision.

Table 3 reports the audit trail linking each Round 1 concern raised by experts to the corresponding revision decision and final competency, consolidating the narrative above into a single reference point.

Round 2: definitional validation and importance ranking

Round 2 pursued three objectives: validation of the 12 refined competency definitions, ambiguity re-assessment of the 23 retained competencies, and importance ranking of all 35 competencies. Two distinct patterns of exclusion emerged: definitional ambiguity precluded four competencies, and insufficient strategic importance consensus precluded five others.

Clarity and ambiguity assessment

Ten of the 12 refined competencies (83.3%) met all clarity thresholds (see Table 4). Four achieved unanimous agreement (100%): work-life integration (M = 4.38), business acumen (M = 4.46), career proactivity (M = 4.54), and virtual collaboration (M = 4.46). Digital agility (M = 3.92, CV = 28.4%, agreement = 76.9%) and human-technology collaboration (M = 3.85, agreement = 69.2%) failed to meet the required thresholds and were excluded.

Twenty-one of the 23 retained competencies maintained satisfactory clarity. Two fell below the clarity mean threshold of 4.0: artificial intelligence (M = 3.92) and decision-making (M = 3.85). Despite high importance ratings, both artificial intelligence (M = 6.00) and decision-making (M = 6.15) were removed from the framework, as the established methodology required that strategic relevance not compensate for definitional ambiguity. Experts indicated that artificial intelligence remained conceptually unclear as an individual-level career competency, whilst decision-making was considered too broad in scope without a defined context or methodology.

Importance rankings

The importance rankings revealed a clear three-tier hierarchy among the 35 retained competencies (see Table 5). The four highest-ranked competencies, namely career adaptability, ethical behavior, digital literacy, and continuous learning, each achieved a mean of 6.31 and a median of 7.00, indicating that at least half of the panel rated these as extremely important. The second-highest cluster (M = 6.23) comprised resilience, complex problem-solving, critical thinking, and digital agility; however, digital agility was subsequently removed owing to clarity failure.

Five competencies were removed owing to insufficient strategic importance consensus: professional social media (M = 4.69, frequency agreement = 62%), and goal-setting, market intelligence, inclusion, and networking (all frequency agreement = 77%). Whilst the latter four approached or exceeded the 5.5 mean threshold, all failed to reach the required 85% frequency agreement criterion, indicating insufficient expert consensus on their strategic essentiality in 4IR contexts.

Competencies excluded from the final framework

Across both rounds in Table 6, 13 competencies were excluded from the final framework. Four competencies were removed at Round 1 on the basis of poor agreement and high variability. A further nine were excluded at Round 2: four owing to definitional ambiguity (artificial intelligence, decision-making, digital agility, and human-technology collaboration) and five owing to failure to achieve the strategic importance frequency agreement threshold (professional social media, goal-setting, market intelligence, inclusion, and networking). The removal of artificial intelligence and decision-making, despite their high importance ratings, illustrates a key methodological principle of the dual-threshold Delphi approach: a competency that cannot be consistently interpreted cannot be reliably assessed or developed within a practical framework.

The final 4IR career competency framework

Following two rounds of expert consensus validation, 26 career competencies met all retention criteria and were included in the final 4IR Career Competency Framework. Table 7 presents the full framework, organized by tier, with validated definitions for each competency. Tier 1 comprises 12 critical core competencies (importance means of 6.00–6.31; frequency agreement of 85% or greater). Tier 2 comprises 14 important contributors (importance means of 5.54–5.92; frequency agreement of 85% or greater).

Sixteen competencies were retained with their original Phase 1 definitions; 10 were refined based on expert feedback. Career adaptability was renamed from adaptability to reflect the career-specific focus established through the Delphi process. Virtual collaboration was developed from the original virtual and online remote work self-efficacy construct, which was removed in its original form but reconceptualised as a distinct collaborative competency during qualitative refinement. Career proactivity and process improvement skills were both derived from the single original proactivity competency, and business acumen incorporated the previously distinct contextual business intelligence construct.

It is worth distinguishing this domain-level classification from the Tier 1/Tier 2 classification reported above. Tier placement was determined quantitatively, by the pre-specified importance-mean and frequency-agreement thresholds applied uniformly to all 26 competencies (see Table 4), and required no researcher judgement. The four domains described below were arrived at differently: they were created afterwards, by researcher judgement, through thematic grouping of the competency definitions rather than through a further round of expert rating or a statistical threshold. The competencies span four broad domains. predetermined categories were imposed on the 26 validated competencies. Instead, the two researchers independently grouped each competency definition according to conceptual similarity in content and focus, compared groupings, and resolved disagreements through discussion until consensus on domain membership was reached. This process is therefore best characterized as researcher judgement informed by, but not tested against, the ICF, Career Construction Theory, and Career Sustainability Theory reviewed earlier. No statistical procedure, such as exploratory or confirmatory factor analysis, was applied to the resulting groupings: adaptive and learning capabilities (career adaptability, continuous learning, resilience, open-mindedness, career proactivity); higher-order cognitive skills (critical thinking, complex problem-solving, creativity, information processing, data analysis, cognitive load management); interpersonal and collaborative skills (collaboration, virtual collaboration, communication skills, teamwork, empathy, cultural awareness, diversity); and professional and operational competencies (ethical behavior, digital literacy, work-life integration, boundary-setting, innovation implementation, process improvement skills, cross-functional skills, business acumen). Consistent with this inductive, non-validated status, the domain structure was not subjected to a separate formal validation step; accordingly, the four-domain architecture should be regarded as a theoretically informed organizational heuristic pending future structural validation through methods such as confirmatory factor analysis or independent expert review of the domain groupings.

Discussion

The study aimed to develop and validate a career competency framework for the Fourth Industrial Revolution tailored to formal-sector professionals navigating 4IR career transitions in South Africa. Framed against South African labor-market inequality, digital exclusion, occupational skills mismatch, and the incomplete professionalization of career development practice the research addressed three objectives: (1) refining the list of career competencies and building consensus on their relevance, (2) assessing and prioritizing the relative importance of the validated competencies for long-term career sustainability, and (3) developing a finalized, expert-validated framework of 4IR career competencies categorized into coherent domains. The primary contribution is a validated framework comprising 26 career competencies organized into four coherent domains, offering individuals, organizations, educators, and policymakers a theoretically grounded and practically applicable tool for navigating career development amid digital and automated transformation.

Research objective 1: refining competencies and establishing consensus

The Delphi process achieved strong panel retention (76.5%), consistent with recommended standards for homogeneous expert groups (Diamond et al., 2014). The framework was reduced from 38 competencies to 35 after Round 1, with 42% of competencies affected through refinement, removal, or the addition of new constructs. Only 21 competencies, those with 90% or greater agreement and low variability (SD < 0.5)—achieved unambiguous consensus, confirming that Delphi consensus measurement requires multiple statistical indicators rather than threshold agreement alone (Beiderbeck et al., 2021; Niederberger and Spranger, 2020).

The disagreement patterns revealed four sources: definitional inadequacy, conceptual overlap, scope boundary disputes, and construct categorization challenges. The elimination of emotional intelligence is particularly noteworthy. Whilst experts identified it as an important career competency for the 4IR, they rejected its operationalisation, citing that it significantly underrepresented its theoretical breadth. This reflects the tension identified by Megahed (2018) and Whiddett and Hollyforde (2003), who argue that competency constructs are context-dependent, and that generic frameworks adapted to specific work contexts risk producing constructs that no longer capture their theoretical meaning. The finding suggests that the operationalisation of emotional intelligence as presented in this study may not have captured the culturally specific relational and communal dimensions that shape its expression in South African professional contexts. The extent to which this rejection reflects broader patterns related to indigenous relational ethics and community-oriented professional practice remains an empirical question that warrants dedicated investigation drawing on African-centered psychology and ubuntu ethics scholarship (e.g., Mangaliso, 2001; Nussbaum, 2003). This is identified as a priority for future research. These additional career competencies confirm that the Delphi validation process can generate diverse ideas and solutions to a problem beyond simple endorsement or rejection (Niederberger and Spranger, 2020).

Research objective 2: prioritizing competencies for career sustainability

The importance rankings produced a clear three-tier hierarchy. The four highest-ranked competencies: career adaptability (M = 6.31), ethical behavior (M = 6.31), digital literacy (M = 6.31), and continuous learning (M = 6.31) reflect priorities that are simultaneously aligned with global frameworks and inflected by South African realities. These findings align with Career Construction Theory (Savickas, 2013, 2002, 2021), which emphasizes adaptive meaning-making over linear career progression, and with the Intelligent Career Framework (Arthur et al., 1995), in which continuous learning corresponds to “knowing why”, adaptability to “knowing how”, and digital literacy to the contemporary instantiation of “knowing what.”

The top-tier rating for ethical behavior (M = 6.31) is especially significant, as ethics typically occupies a secondary position in traditional competency frameworks. It challenges traditional assumptions and suggests that, in digitally mediated 4IR workplaces, ethical behavior is a core competency rather than an optional addition, especially as most work occurs digitally. This finding is consistent with the growing international literature on digital ethics and responsible AI use in organizations (Jobin et al., 2019; Taddeo and Floridi, 2018), suggesting that South African expert opinion converges with broader concerns about trust and accountability in automated work environments.

The removal of goal-setting (M = 5.31) and the moderate-but-excluded rating of networking (M = 5.69) are the most counterintuitive findings, challenging established career management theories that historically placed significant emphasis on both goal-directed behavior and social capital accumulation (Batistic and Tymon, 2017; Hall et al., 2018). In a South African labor market characterized by 32.7% unemployment and frequent involuntary career disruption, fixed goal-directedness may function less as a strategic asset and more as a source of frustration when structural barriers prevent goal attainment. Broad adaptive capacity is therefore a more viable orientation than specific goal pursuit. Similarly, networking may carry less weight where AI-driven recruitment tools increasingly substitute for relationship-based access, and where geographical and infrastructural inequalities limit the reach of professional networks outside major urban centers such as Gauteng and Cape Town. The findings also demonstrate that South African expert opinion aligns substantially with global consensus despite different socio-economic structures. The prioritized competencies mirror the World Economic Forum's (2025) top-ranked workforce capabilities, namely analytical thinking, creativity, active learning, and critical thinking, suggesting that the convergence between locally grounded validation and international priorities reflects the structural demands that digital transformation imposes across diverse economies. What is specifically South African about the final framework is therefore not merely its site of development but its prioritization pattern: the exclusion of networking and goal-setting, and the elevated standing of ethical behavior and adaptive capacities, reflect local labor-market structural barriers, AI-driven recruitment exclusion risks, and digital infrastructure constraints rather than a generic restatement of international competency lists.

Research objective 3: the final framework and its domains

The final framework's organization into 26 competencies across four domains reflects a coherent theoretical architecture. Its most distinctive feature is the strong representation of interpersonal competencies (six competencies in the interpersonal and collaborative domain), which challenges the widespread assumption that 4IR environments prioritize technical over human-centered capabilities. This is consistent with the Operator 4.0 framework (Romero et al., 2016), which proposes that as human-technology collaboration intensifies, the demand for interpersonal skills increases rather than diminishes, a pattern the South African expert panel clearly endorsed.

Career competencies, employability competencies, and job performance competencies

A theoretical tension within the framework warrants acknowledgment: several competencies, particularly process improvement skills, innovation implementation and business acumen, sit more naturally within a job performance competency category than within the career self-management orientation that characterizes the Intelligent Career Framework, Career Construction Theory, and the Akkermans et al. (2013) career competencies model. The traditional distinction positions career competencies as capabilities individuals deploy to navigate, sustain, and advance their own careers, whereas job performance competencies are behaviors and capabilities valued and assessed by employers. We argue that the 4IR context substantially collapses this distinction. In a labor market characterized by rapid technological change, organizational restructuring, and the increasing permeability of occupational boundaries, the ability to understand organizational strategy (business acumen) and to improve operational processes (process improvement skills) functions not merely as a job performance requirement but as a career navigation capability. Innovation implementation follows the same logic, the capacity to turn ideas into practical solutions determines whether an individual is positioned as a source of value during organizational restructuring or displaced by it, which makes it a career-survival capability rather than only a performance metric. Workers who cannot read the strategic direction of their organizations, implement innovations, or contribute to process optimisation are less able to anticipate career threats, identify developmental priorities, or position themselves for transitions. Expert consensus on the inclusion ofbusiness acumen, process improvement skills, and innovation implementation within a career framework suggests that South African practitioners recognize this blurring of traditional categories as a structural feature of 4IR employment, rather than a conceptual error. Nevertheless, we acknowledge that the framework is hybrid in character, and future research should examine whether career-oriented and performance-oriented competencies function differently in predicting career outcomes vs. job performance outcomes.

The deliberate restriction of technology and data domains to two competencies merits explanation. The exclusion of three technology-related competencies namely, artificial intelligence, digital agility, and human-technology collaboration that despite their high importance ratings demonstrates the perceived relevance cannot compensate for operational ambiguity. This makes a methodological contribution to debates on framework legitimacy (Francis-Smythe et al., 2012; Whiddett and Hollyforde, 2003): in a South African context where digital skill definitions vary widely across sectors and infrastructure levels, prescribing competencies that lack definitional consensus would undermine rather than support workforce development. The rapidly evolving nature of AI competencies in particular suggests that premature formalization risks producing frameworks that are obsolete before implementation, a risk that the Delphi clarity thresholds appropriately guarded against. For curriculum designers, this finding has direct implications. The exclusion of these competencies despite near-unanimous importance endorsement indicates that they cannot yet be assessed against consensus-validated definitions, suggesting they may be more effectively developed through embedded application within existing modules rather than as standalone assessed outcomes. Curriculum revision cycles for AI-related content will also need to be substantially shorter than for the stable competencies validated in this framework.

The inclusion of work-life integration (Tier 1) and boundary-setting (Tier 2) integrates wellbeing directly into the framework's structure rather than treating it as peripheral. This aligns with Career Construction Theory's assumptions about career-life integration (Savickas, 2013; Kundi et al., 2021) and reflects the specific pressures of South African professionals managing 24-h connectivity in contexts where domestic and professional boundaries are frequently contested by infrastructure limitations and informal employment arrangements. The simultaneous retention of work-life integration (Tier 1) and boundary-setting (Tier 2) presents an apparent conceptual tension that warrants direct engagement. The boundary management literature distinguishes clearly between integration and segmentation as competing strategies for managing the work-life interface (Nippert-Eng, 2008; Kreiner et al., 2009), and individuals differ systematically in their preferences for boundary permeability. Work-life integration, as defined in this framework, involves creating flexible boundaries allowing work and personal activities to complement each other, whereas boundary-setting involves creating clear separation between work and personal time despite continuous digital connectivity. We argue that the 4IR context requires individuals to hold both capabilities simultaneously, not as contradictions but as contextually deployed strategies. Integration is the appropriate orientation when technology enables productive blending of professional and personal demands; boundary-setting is necessary when 24-h digital connectivity threatens cognitive depletion and recovery. The expert panel's consensus on both competencies supports the interpretation that South African practitioners recognize the dual demand: professionals in digitally mediated work environments must be capable of both integrating flexibly and separating deliberately, depending on the situational requirements. This is consistent with recent work on boundary management flexibility, which argues that the ability to transition between integration and segmentation strategies is itself a career capability (Kossek and Lautsch, 2012).

Practical implications

Practical implications are organized below by stakeholder group, since Tier 1 and Tier 2 competencies serve a different function for each (Table 8). For career counselors, in resource-constrained settings, such as township-based career counseling centers or public sector skills development units, the framework's tier architecture, organized here according to the provisional four-domain heuristic described above rather than an independently validated structure, provides a triage principle: Tier 1 competencies should be assessed and developed first, establishing the foundational adaptive, ethical, and digital capabilities without which Tier 2 development is unlikely to be effective. Practically, a career counselor working with a client entering the labor market would begin with career adaptability, digital literacy, and communication skills before progressing to creativity, data analysis, or business acumen. A practical readiness indicator is whether a client can independently sustain Tier 1 capabilities, such as managing change without close supervision or using digital tools competently, before Tier 2 development, such as cross-functional or strategic competencies, is introduced.

For educators and curriculum designers, the implication is that competency development should be sequenced rather than simultaneous, with Tier 1 competencies embedded in induction and foundation phase curricula and Tier 2 competencies integrated into mid-career or later stage development and specialization tracks. Among Tier 1 competencies, digital literacy and communication skills lend themselves to structured instructional delivery, whereas career adaptability and resilience are more effectively developed through work-integrated learning. This suggests that teachable competencies can be front-loaded in curricula, whilst experiential ones require sustained workplace engagement across the programme duration. As noted above, AI-related competencies were excluded from the validated framework owing to definitional ambiguity; educators should therefore treat these as embedded, application-based learning outcomes revisited each curriculum cycle, rather than as standalone assessed competencies benchmarked against the stable Tier 1 and Tier 2 definitions.

For HR practitioners, Tier 1 competencies are best positioned as non-negotiable criteria in recruitment screening, onboarding, and probation review, given their near-unanimous expert endorsement on both clarity and strategic importance. Tier 2 competencies are better suited to performance development plans, promotion readiness assessments, and specialist role profiling, where their status as important but non-universal contributors allow for role-specific weighting rather than uniform application across all positions.

For policymakers, the framework offers an evidence base for aligning national skills planning with locally validated, rather than imported, competency priorities. Tier 1 competencies, particularly digital literacy, ethical behavior, and career adaptability, provide a defensible minimum content standard for publicly funded 4IR reskilling and active labor market programmes, while the exclusion of goal-setting and networking from the final framework signals that programme design premised on individual goal-directedness or network-building alone is unlikely to address the structural barriers described earlier.

Across all four groups, the framework also interfaces with South Africa's existing skills development architecture. The National Qualifications Framework (NQF) and the Skills Development Act (No. 97 of 1998) currently prioritize occupation-specific unit standards; the present framework offers a complementary layer of transferable career competencies that could inform the design of cross-occupational learning programmes at NQF Levels 5–7, where the transition from technical training to professional practice is most critical for 4IR readiness.

Limitations

Several limitations should be considered when interpreting these findings. These limitations span four areas: sample composition and generalisability, methodological design choices, measurement of consensus, and the provisional status of the domain structure. The reliance on purposive sampling means the findings reflect the perspectives of a selected group of South African professionals within specific sectors and roles. The Delphi method's structured format, while mitigating groupthink, also constrains the depth of deliberate exchange possible in an iterative group process. The decision to conclude after two rounds may have left some competencies, particularly those reflecting authentic contextual variation, less stabilized further with additional iterations. We acknowledge, however, that a third round particularly for refined and newly created competencies would have provided additional confidence in stability, and we note this as a limitation. The expert panel represents the formal, professional sector of the South African economy, with expertise concentrated in industrial psychology, human resource management, and education. The study's expert panel and primary application logic are oriented toward career psychology and human resource practice rather than educational programme design. Nevertheless, the framework offers curriculum designers an empirically grounded alternative to ad hoc competency lists derived from employer surveys or policy documents, and the finding that AI-related competencies resist stable operationalisation has direct implications for how higher education institutions approach the formalization of technology-related learning outcomes. No panel member was drawn from construction, mining, agriculture, retail, public health, or the informal economy, which collectively account for the majority of South African employment. The framework's applicability to these sectors and to workers without postgraduate qualifications should not be assumed and requires further validation. A further limitation is the absence of a formal inter-rater reliability measure, such as Krippendorff's alpha or Kendall's coefficient of concordance, to quantify the degree of expert agreement beyond descriptive statistics and percentage agreement figures, the changing item set between Round 1 and Round 2 made a single such coefficient difficult to apply meaningfully across the whole study, which is why per-round, per-item triangulated thresholds were used instead. Whilst the second author's examination of the qualitative analysis provided a form of secondary review, this does not substitute for a formal reliability coefficient. A measure such as Kendall's coefficient of concordance (W), which is well suited to ranked Delphi importance data, collected on a stable item set, would have been particularly appropriate in a design with a fixed competency list across rounds. Future Delphi studies in this domain should incorporate such measures to enable readers to assess the strength of consensus independently of the authors' interpretive framing.

Despite evidence of alignment with international frameworks, the South African context may limit the direct applicability of these findings to other national or organizational settings, especially those with varying technological infrastructure or cultural dynamics. Moreover, rapidly evolving 4IR technologies, particularly AI, may render specific competency definitions less stable over time, necessitating periodic revisions of the current framework.

Recommendations for future research

Several directions for future research emerge from this study. First, future studies should expand both sample size and geographic scope to include participants from all South African provinces and from a broader range of occupational levels, including those with limited formal qualifications or in precarious employment to strengthen representativeness and improve applicability to sectors and qualification levels beyond the formal professional sector examined here. Second, quantitative and mixed-methods research should be pursued to validate the four-domain competency framework identified in this study. Survey-based approaches could test the framework across large, representative samples, enabling statistical examination of how competency priorities vary by sector, educational level, age group, and region, moving the framework from an expert-derived construct to an empirically validated instrument with population-level generalisability. Third, longitudinal research is needed to track how 4IR competency requirements evolve as technologies such as generative artificial intelligence, automation, and platform-mediated work continue to mature. Comparative studies across African economies at different stages of 4IR adoption would contribute to more contextually nuanced theoretical frameworks that move beyond the Western applicability assumptions currently dominating the field.

Conclusion

The validated framework of 26 career competencies, organized into two tiers and four domains, provides the South African career development field with an empirically grounded foundation that has, until now, been absent. The framework's principal methodological contribution lies in the dual-threshold validation principle, which required every retained competency to demonstrate both definitional precision and strategic importance consensus. This principle, illustrated most clearly by the exclusion of artificial intelligence and decision-making despite their high importance ratings, establishes that perceived relevance cannot substitute for operational clarity in frameworks intended for assessment and development use.

Theoretically, the study contributes to ongoing debates about the relationship between career competencies and job performance competencies, the operationalisation of established constructs such as emotional intelligence in culturally specific contexts, and the simultaneous demand for integration and boundary-setting in digitally mediated professional environments. The framework's strong representation of interpersonal and adaptive competencies, rather than primarily technical capabilities, challenges the assumption that 4IR readiness is fundamentally a digital skills problem and aligns with emerging international evidence that human-centered capabilities remain essential as technology integration intensifies.

Practically, the two-tier structure provides a sequencing logic for career psychologists, HR practitioners, educators, and policymakers working within resource constraints: Tier 1 competencies should anchor developmental interventions before Tier 2 capabilities are pursued. The framework's derivation from local expert consensus across multiple sectors and provinces ensures contextual relevance to South African skills development priorities, whilst its partial convergence with World Economic Forum (2025) priorities suggests potential applicability to comparable emerging economy contexts.

Several limitations qualify these contributions. The expert panel represents the formal professional sector, and applicability to the informal economy, primary industries, and workers without postgraduate qualifications requires further investigation. The four-domain architecture was derived inductively and has not been subjected to independent structural validation. The absence of a formal inter-rater reliability measure constrains the extent to which the strength of consensus can be independently evaluated. Finally, the rapidly evolving nature of 4IR technologies, particularly generative artificial intelligence, means that periodic revision of both the competency definitions and the framework's scope will be necessary. The framework should be understood as a validated and defensible starting point for South African career competency development, not as a static endpoint.

References

• 1

  AkkermansJ.BrenninkmeijerV.HuibersM.BlonkR. W. B. (2013). Competencies for the contemporary career. J. Career Dev.40, 245–267. doi: 10.1177/0894845312467501

  • CrossRef
  • Google Scholar

• 2

  AkkermansJ.KubaschS. (2017). “#Trending topics in careers: a review and future research agenda,” in Career Development International, Vol. 22, (Bingley: Emerald Group Publishing Ltd.), 586–627. doi: 10.1108/CDI-08-2017-0143

  • CrossRef
  • Google Scholar

• 3

  ArthurM. B.ClamanP. H.DeFillippiR. J. (1995). Intelligent enterprise, intelligent careers. Acad. Manag. Exec.9, 7–22. doi: 10.5465/ame.1995.9512032185

  • CrossRef
  • Google Scholar

• 4

  AvellaJ. R. (2016). “Delphi panels: research design, procedures, advantages, and challenges,” in International Journal of Doctoral Studies, Vol. 11, doi: 10.28945/3561 (Accessed September 15, 2025).

  • CrossRef
  • Google Scholar

• 5

  BalP. M. (2024). “Sustainable careers,” in Concise Encyclopedia of Corporate Social Responsibility (Cheltenham: Edward Elgar Publishing Ltd.), 192–195. doi: 10.4337/9781800880344.ch36

  • CrossRef
  • Google Scholar

• 6

  BarnesN.du PlessisM.FrantzJ. (2022). Career competencies for academic career progression: experiences of academics at a South African university. Front. Educ.7, 1–11. doi: 10.3389/feduc.2022.814842

  • CrossRef
  • Google Scholar

• 7

  BatisticS.TymonA. (2017). Networking behaviour, graduate employability: a social capital perspective. Educ. Train.59, 374–388. doi: 10.1108/ET-06-2016-0100

  • CrossRef
  • Google Scholar

• 8

  BattA.WilliamsB.RichJ.TavaresW. (2021). A six-step model for developing competency frameworks in the healthcare professions. Front. Med.8, 1–9. doi: 10.3389/fmed.2021.789828

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9

  BaulsV. A.TuroffM. (2011). Scenario construction via Delphi and cross-impact analysis. Technol. Forecast. Soc. Change78, 1579–1602. doi: 10.1016/j.techfore.2011.03.014

  • CrossRef
  • Google Scholar

• 10

  BeiderbeckD.FrevelN.von der GrachtH. A.SchmidtS. L.SchweitzerV. M. (2021). The impact of COVID-19 on the European football ecosystem—A Delphi-based scenario analysis. Technol. Forecast. Soc. Change165, 1–20. doi: 10.1016/j.techfore.2021.120577

  • CrossRef
  • Google Scholar

• 11

  BeigiM.ShimohammadiM.ArthurM. (2018). Intelligent career success the case of distinguished academics.pdf. J. Vocat. Behav.107, 261–275. doi: 10.1016/j.jvb.2018.05.007

  • CrossRef
  • Google Scholar

• 12

  BeltonI.MacdonaldA.WrightG.HamlinI. (2019). Improving the practical application of the Delphi method in group-based judgement: a six-step prescription for a well-founded and defensible process. Technol. Forecast. Soc. Change147, 72–82. doi: 10.1016/j.techfore.2019.07.002

  • CrossRef
  • Google Scholar

• 13

  BeltonI.WrightG.SissonsA.BolgerF.CrawfordM. M.HamlinI.et al. (2021). Delphi with feedback of rationales: how large can a Delphi group be such that participants are not overloaded, de-motivated, or disengaged?Technol. Forecast. Soc. Change170, 1–17. doi: 10.1016/j.techfore.2021.120897

  • CrossRef
  • Google Scholar

• 14

  Blasco-BlascoO.Pinto-DelacadenaP. A.Vinueza-CabezasA.LiernV. (2025). Crafting a competency-based questionnaire based on fuzzy Delphi: a proposal for enhanced hiring process. Qual. Quant. 59, 3245–3269. doi: 10.1007/s11135-025-02086-x

  • CrossRef
  • Google Scholar

• 15

  BonaccorsiA.ApredaR.FantoniG. (2020). Expert biases in technology foresight. Why they are a problem and how to mitigate them. Technol. Forecast. Soc. Change151, 1–17. doi: 10.1016/j.techfore.2019.119855

  • CrossRef
  • Google Scholar

• 16

  BouwmansM.LubX.OrlowskiM.NguyenT. V. (2024). Developing the digital transformation skills framework: a systematic literature review approach. PLoS ONE 19 doi: 10.1371/journal.pone.0304127

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  BoyatzisR. (1982). The Competent Manager: A Model for Effective Performance. Available online at: https://www.semanticscholar.org/paper/aa283f3ebcc00458955138c62793dd7f9dc9fb48

  • Google Scholar

• 18

  BradyS. R. (2015). Utilizing and adapting the Delphi method for use in qualitative research. Int. J. Qual. Methods 14. doi: 10.1177/1609406915621381

  • CrossRef
  • Google Scholar

• 19

  BraunV.ClarkeV. (2006). Using thematic analysis in psychology. Qual. Res. Psychol.3, 77–101. doi: 10.1191/1478088706qp063oa

  • CrossRef
  • Google Scholar

• 20

  CampionM.FinkA. A.RuggebergB. J.CarrL. L.PhillipsG. M.OdmanR. B. (2011). Doing competencies well: best practices in competency modeling. Pers. Psychol.64, 225–262. doi: 10.1111/j.1744-6570.2010.01207.x

  • CrossRef
  • Google Scholar

• 21

  ChakaC. (2020). Skills, competencies and literacies attributed to 4IR/Industry 4.0: scoping review. J. Int. Fed. Libr. Assoc. Inst.46, 369–399. doi: 10.1177/0340035219896376

  • CrossRef
  • Google Scholar

• 22

  CoetzeeM.SchreuderM. D. (2020). Careers: An Organisational Perspective (6th edn.). Juta and Co, Ltd. Available online at: https://ebookcentral.proquest.com/lib/ujlink-ebooks/reader.action?docID=6634864andppg=4 (Accessed August 22, 2025).

  • Google Scholar

• 23

  Crowley-HenryC.AlmeidaS.BertoneS.GunasekaraA. (2023). The multilevel intelligent career framework an exploration and application to skilled migrants. Career Dev. Int.28, 474–499. doi: 10.1108/CDI-04-2022-0097

  • CrossRef
  • Google Scholar

• 24

  DalkeyN.HelmerO. (1962). An Experimental Application of the Delphi Method to the Use of Experts | Enhanced Reader. United States Air Force under Project RAND, (July). Available online at: https://www.rand.org/content/dam/rand/pubs/research_memoranda/2009/RM727.1.pdf

  • Google Scholar

• 25

  De VosA.Van der HeijdenB. I. J. M.AkkermansJ. (2020). Sustainable careers: towards a conceptual model. J. Vocat. Behav.117:103196. doi: 10.1016/j.jvb.2018.06.011

  • CrossRef
  • Google Scholar

• 26

  DeFillippiR. J.ArthurM. B. (1994). The boundaryless career: a competency-based perspective. J. Organ. Behav.15, 307–324.

  • Google Scholar

• 27

  DelbecqA. L.Van de VenA.GustafsonD. H. (1975). Group Techniques for Program Planning: A Guide to Nominal Group and Delphi Processes. Glenview, IL: Scott Foresman and Company.

  • Google Scholar

• 28

  Deloitte (2019). Skills Change, but Capabilities Endure. The Future of Work and Importance of Human Capabilities. New York, NY: Deloitte Insights.

  • Google Scholar

• 29

  Department Higher Education and Training (2014). Employer Guideline to OFO Codes in the Fasset Sector. Available online at: https://www.fasset.org.za/downloads/2016 OFO Guideline V3Updated.pdf (Accessed June 9, 2023).

  • Google Scholar

• 30

  DiamondI. R.GrantR. C.FeldmanB. M.PencharzP. B.LingS. C.MooreA. M.et al. (2014). Defining consensus: a systematic review recommends methodologic criteria for reporting of Delphi studies. J. Clin. Epidemiolo.67, 401–409. doi: 10.1016/j.jclinepi.2013.12.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  EinarsdóttirS.RoundsJ.SuR. (2010). Holland in iceland revisited: an emic approach to evaluating U.S. vocational interest models. J. Counsel. Psychol.57, 361–367. doi: 10.1037/a0019685

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  Francis-SmytheJ.HaaseS.ThomasE.SteeleC. (2012). Development and validation of the career competencies indicator (CCI). J. Career Assess.21, 227–248. doi: 10.1177/1069072712466724

  • CrossRef
  • Google Scholar

• 33

  GartoumiK. I.TekouabouS. C. K. (2024). Exploring African skills 4.0 required for 4IR industry: a systematic literature review. Data Policy6, e36-1–e36-23. doi: 10.1017/dap.2024.28

  • CrossRef
  • Google Scholar

• 34

  GeorgeA. S.BaskarT.KarthikeyanM. M. (2025). Future of work trends: Key insights for 2025. Partners Univ. Multidiscip. Res. J.2, 48–78. doi: 10.5281/zenodo.14738253

  • CrossRef
  • Google Scholar

• 35

  GoneseD.NgepahN. N. (2024). Fourth Industrial Revolution technologies and sectoral employment in South Africa. Cogent Soc. Sci.10, 1–20. doi: 10.1080/23311886.2024.2382282

  • CrossRef
  • Google Scholar

• 36

  HallD. T.YipJ.DoironK. (2018). Protean careers at work: self-direction and values orientation in psychological success. Ann. Rev. Organ. Psychol. Organ. Behav.5, 129–156. doi: 10.1146/annurev-orgpsych-032117-104631

  • CrossRef
  • Google Scholar

• 37

  HassonF.KeeneyS. (2011). Enhancing rigour in the Delphi technique research. Technol. Forecast. Soc. Change, 78, 1695–1704. doi: 10.1016/j.techfore.2011.04.005

  • CrossRef
  • Google Scholar

• 38

  HassonF.KeeneyS.McKennaH. (2025). “Revisiting the Delphi technique—Research thinking and practice: a discussion paper,” in International Journal of Nursing Studies, Vol. 168, (Amsterdam: Elsevier Ltd.) doi: 10.1016/j.ijnurstu.2025.105119

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  HattinghS. (2016). Skills planning for the disruptive Fourth Industrial Revolution. HR Future 1–7. doi: 10.10520/EJC199869

  • CrossRef
  • Google Scholar

• 40

  HirschiA. (2018). The Fourth Industrial Revolution: issues and implications for career research and practice. Career Dev. Q. 66, 192–204. doi: 10.1002/cdq.12142

  • CrossRef
  • Google Scholar

• 41

  HoleyE. A.FeeleyJ. L.DixonJ.WhittakerV. J. (2007). An exploration of the use of simple statistics to measure consensus and stability in Delphi studies. BMC Med. Res. Methodol.7:52. doi: 10.1186/1471-2288-7-52

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  Human Sciences Research Council (2018). Policy options framework for the Fourth Industrial Revolution in South Africa. Pretoria.

  • Google Scholar

• 43

  JoH.ParkM.SongJ. H. (2024). Career competencies: an integrated review of the literature. Eur. J. Train. Dev.48, 805–832. doi: 10.1108/EJTD-04-2023-0052

  • CrossRef
  • Google Scholar

• 44

  JobinA.IencaM.VayenaE. (2019). The global landscape of AI ethics guidelines. Nat. Mach. Intell.1, 389–399. doi: 10.1038/s42256-019-0088-2

  • CrossRef
  • Google Scholar

• 45

  JoshiA.KaleS.ChandelS.PalD. (2015). Likert scale: explored and explained. Br. J. Appl. Sci. Technol. 7, 396–403. doi: 10.9734/BJAST/2015/14975

  • CrossRef
  • Google Scholar

• 46

  KhodyakovD.GrantS.KrogerJ.BaumanM. (2023). RAND Methodological Guidance for Conducting and Critically Appraising Delphi Panels Manual. Available online at: www.rand.org/about/research-integrity (Accessed June 4, 2025).

  • Google Scholar

• 47

  KossekE. E.LautschB. A. (2012). Work–family boundary management styles in organizations: a cross-level model. Organ. Psychol. Rev.2, 152–171. doi: 10.1177/2041386611436264

  • CrossRef
  • Google Scholar

• 48

  KreinerG. E.HollensbeE. C.SheepM. L. (2009). Balancing borders and bridges: negotiating the work-home interface via boundary work tactics. Acad. Manag. J.52, 704–730. doi: 10.5465/amj.2009.43669916

  • CrossRef
  • Google Scholar

• 49

  KuijpersM.ScheerensJ. (2006). Career competencies for the modern career. J. Career Dev.32, 303–319. doi: 10.1177/0894845305283006

  • CrossRef
  • Google Scholar

• 50

  KundiY. M.Hollet-HaudebertS.PetersonJ. (2021). Linking protean and boundaryless career attitudes to subjective career success: a serial mediation model. J. Career Assess.29, 263–282. doi: 10.1177/1069072720959782

  • CrossRef
  • Google Scholar

• 51

  LaherS.CockcroftK. (2018). “Psychological assessment in South Africa. Research and applications,” in Psychological Assessment in South Africa (Johannesburg: Wits University Press).

  • Google Scholar

• 52

  LeiY.-W.KimR. (2025). Automation and augmentation: artificial intelligence, robots, and Work. 14:39. doi: 10.1146/annurev-soc-090523-050708

  • CrossRef
  • Google Scholar

• 53

  LinstoneH. A.TuroffM.HelmerO. (2002). The Delphi Method: Techniques and Applications. Newark, NJ.

  • Google Scholar

• 54

  MangalisoM. P. (2001). Building competitive advantage from ubuntu: management lessons from South Africa. Acad. Manag. Perspect.15, 23–33 doi: 10.5465/ame.2001.5229453

  • CrossRef
  • Google Scholar

• 55

  MarchalE.MelletK.RieucauG. (2007). Job board toolkits: internet matchmaking and changes in job advertisements. Hum. Relat.60, 1091–1113. doi: 10.1177/0018726707081159

  • CrossRef
  • Google Scholar

• 56

  MareeJ. G. (2009). Career counselling in the 21st century: South African institutions of higher education at the crossroads. S. Afr. J. High. Educ.23, 429–435. doi: 10.10520/EJC37546

  • CrossRef
  • Google Scholar

• 57

  McClellandD. C. (1973). Testing for competence rather than for “intelligence.”Am. Psychol.28, 1–14. doi: 10.1037/h0034092

  • Google Scholar

• 58

  McMahonM.AbkhezrP. (2025). Career adaptability and career resilience: a systems perspective. Int. J. Educ. Vocat. Guid. doi: 10.1007/s10775-025-09739-1

  • CrossRef
  • Google Scholar

• 59

  McMillanS. S.KingM.TullyM. P. (2016). How to use the nominal group and Delphi techniques. Int. J. Clin. Pharm.38, 655–662. doi: 10.1007/s11096-016-0257-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60

  MegahedN. (2018). A critical review of the literature and practice of competency modelling. KnE Soc. Sci.3:104. doi: 10.18502/kss.v3i10.3106

  • CrossRef
  • Google Scholar

• 61

  MtshaliS.DhanpatN. (2026). Establishing career competencies for the Fourth Industrial Revolution: a qualitative study of expert perspectives in South Africa. Front. Psychol.17:1823256. doi: 10.3389/fpsyg.2026.1823256

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  MukherjeeN.HugéJ.SutherlandW. J.McneillJ.Van OpstalM.Dahdouh-GuebasF.et al. (2015). The Delphi technique in ecology and biological conservation: applications and guidelines. Methods Ecol. Evol.6, 1097–1109. doi: 10.1111/2041-210X.12387

  • CrossRef
  • Google Scholar

• 63

  MunsamyM.DhanpatN.BarkhuizenE. N. (2023). The development and validation of a digital leadership competency scale. Acta Commer.23, 1–15. doi: 10.4102/ac.v23i1.1057

  • CrossRef
  • Google Scholar

• 64

  NaidooA. V.PretoriusT. B.NicholasL. J. (2024). “The emergence of career psychology in South Africa: a socio-historical perspective,” in The emergence of Career Psychology in South Africa: A socio-his Career psychology in the South African context 4th edn. eds. G. B. Stead and M. B. Watson (Pretoria: J. L. van Schaik), 1–20.

  • Google Scholar

• 65

  NasaP.JainR.JunejaD. (2021). Delphi methodology in healthcare research: how to decide its appropriateness. World J. Methodol.11, 116–129. doi: 10.5662/wjm.v11.i4.116

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  National Planning Commission (2012). Our Future : Make It Work: National Development Plan, 2030. Pretoria: National Planning Commission.

  • Google Scholar

• 67

  Ndung'uN.SignéL. (2020). The Fourth Industrial Revolution and digitization will transform Africa into a global powerhouse. Foresight Africa 2020 Report, 60–73. Available online at: https://www.brookings.edu/research/the-fourth-industrial-revolution-and-digitization-will-transform-africa-into-a-global-powerhouse/%0Ahttps://www.itu.int/en/ITU-D/Statistics/Pages/publications/mis2017/methodology.aspx (Accessed February 5, 2024).

  • Google Scholar

• 68

  NeamtuD. M.BejinaruR.AnichitiA.ButnaruG. I.HapenciucC. V. (2025). Current status and perspectives regarding the skills and meta-competencies needed by employees in the era 4.0. Kybernetes. 55, 3028–3051. doi: 10.1108/K-10-2024-2918

  • CrossRef
  • Google Scholar

• 69

  NiederbergerM.KöberichS. (2021). Coming to consensus: the Delphi technique. Eur. J. Cardiovasc. Nurs.20, 692–695. doi: 10.1093/eurjcn/zvab059

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  NiederbergerM.SchifanoJ.DeckertS.HirtJ.HombergA.KöberichS.et al. (2024). Delphi studies in social and health sciences—Recommendations for an interdisciplinary standardized reporting (DELPHISTAR). Results of a Delphi study. PLoS ONE19:e0304651. doi: 10.1371/journal.pone.0304651

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71

  NiederbergerM.SprangerJ. (2020). Delphi technique in health sciences: a map. Front. Public Health8, 1–10. doi: 10.3389/fpubh.2020.00457

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  Nippert-EngC. E. (2008). Home and Work: Negotiating Boundaries Through Everyday Life. Chicago, IL: University of Chicago Press.

  • Google Scholar

• 73

  NussbaumB. (2003). Ubuntu: reflections of a South African on our common humanity. Reflections4, 21–26. doi: 10.1162/152417303322004175

  • CrossRef
  • Google Scholar

• 74

  OkoliC.PawlowskiS. D. (2004). The Delphi method as a research tool: an example, design considerations and applications. Inf. Manag.42, 15–29. doi: 10.1016/j.im.2003.11.002

  • CrossRef
  • Google Scholar

• 75

  Presidential Commission on the Fourth Industrial Revolution (2020). “Report of the presidential commission on the 4th industrial revolution,” in Presidential Commission on the Fourth Industrial Revolution (Pretori: South African Government).

  • Google Scholar

• 76

  RajhansV.VeR.MohammadC.RegeS. (2021). Development and validation of competency framework for teaching management of refractive errors: a participatory Delphi approach. Indian J. Ophthalmol.69, 3690–3696. doi: 10.4103/ijo.IJO_3312_20

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  RomeroD.BernusP.NoranO.StahreJ.Fast-BerglundÅ. (2016, September). “The operator 4.0: Human cyber-physical systems & adaptive automation towards human-automation symbiosis work systems,” in IFIP international conference on advances in production management systems (Cham: Springer International Publishing), 677–686.

  • Google Scholar

• 78

  SamuelO. M.MoagiT. (2022). The emerging work system and strategy for skills transition in South Africa. Manag. Res. Rev.45, 1503–1523. doi: 10.1108/MRR-05-2021-0356

  • CrossRef
  • Google Scholar

• 79

  SargentL. D.DombergerS. R. (2007). Exploring the development of a protean career orientation: values and image violations. Career Dev. Int.12, 545–564. doi: 10.1108/13620430710822010

  • CrossRef
  • Google Scholar

• 80

  SavickasM. (2013). “Career construction theory and practice,” in Career Development and Counselling: Putting Theory and Research to Work, 2nd edn., eds S. D. Brown and R. W. Lent (John Wiley and Sons), 147–186.

  • Google Scholar

• 81

  SavickasM. (2021). Career Adaptability. Hoboken, NJ: 48HrBooks.

  • Google Scholar

• 82

  SavickasM. L. (2002). “Career construction: a developmental theory of vocational behavior. in Career Choice and Development, 4th edn., ed. D. Brown (Hoboken, NJ: John Wiley and Sons, Inc.), 149–205.

  • Google Scholar

• 83

  SavickasM. L. (2005). The theory and practice of career construction. in Career Development and Counselling: Putting Theory and Research to Work, eds. S. D. Brown and R. W. Lent (Jogn Wiley and Sons, Inc.), 42–70.

  • Google Scholar

• 84

  SchifanoJ.NiederbergerM. (2025). How Delphi studies in the health sciences find consensus: a scoping review. Syst. Rev.14, 1–13. doi: 10.1186/s13643-024-02738-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 85

  SchwabK. (2017). The Fourth Industrial Revolution. New York, NY: Penguin Random House.

  • Google Scholar

• 86

  SekayiD.KennedyA. (2017). Qualitative Delphi method: a four round process with a worked example. Qualit. Rep. doi: 10.46743/2160-3715/2017.2974

  • CrossRef
  • Google Scholar

• 87

  SetiT. M.NgqulanaA.SindesiO. A. (2025). Industry perceptions of employability skills for agricultural graduates in the Fourth Industrial Revolution. SA J. Ind. Psychol.51, 1–11. doi: 10.4102/sajip.v51i0.2214

  • CrossRef
  • Google Scholar

• 88

  SikhakhaneB.LubbeS. (2005). “Preliminaries into problems to access information-the digital divide and rural communities,” in Peer Reviewed Article, Vol. 7 doi: 10.4102/sajim.v7i3.273

  • CrossRef
  • Google Scholar

• 89

  SkulmoskiG. J.HartmanF. T.KrahnJ. (2007). The Delphi method for graduate research. J. Inf. Technol. Educ. Res.6, 1–21.

  • Google Scholar

• 90

  SpencerL. M.SpencerS. M. (1993). “Competence at work,” in CME. Chartered Mechanical Engineer (New York, NY: John Wiley and Sons, Inc.)

  • Google Scholar

• 91

  SpurkD.HirschiA.DriesN. (2019). “Antecedents and outcomes of objective versus subjective career success: competing perspectives and future directions,” in Journal of Management, Vol. 45 (SAGE Publications Inc.), 35–69. doi: 10.1177/0149206318786563

  • CrossRef
  • Google Scholar

• 92

  Statistics South Africa (2026). South Africa's Youth and the Labour Market in Q1 2026. Available online at: https://www.statssa.gov.za/?p=19526 (Accessed January 27, 2026).

  • Google Scholar

• 93

  SultanaR. G. (2024). Four ‘dirty words' in career guidance: from common sense to good sense. Int. J. Educ. Vocat. Guid.24, 1–19. doi: 10.1007/s10775-022-09550-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 94

  TaddeoM.FloridiL. (2018). How AI can be a force for good. Science361, 751–752.

  • Google Scholar

• 95

  TerbraakM.MajorM.JørstadH.Scholte op ReimerW.van der SchaafM. (2024). Home-based cardiac rehabilitation in older adults: expert-recommendations for physiotherapist-led care to improve daily physical functioning and reduce comorbidity-related barriers. Eur. J. Physiother.26, 288–298. doi: 10.1080/21679169.2023.2276712

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 96

  von der GrachtH. A. (2012). Consensus measurement in Delphi studies. Review and implications for future quality assurance. Technol. Forecast. Social Change, 79, 1525–1536. doi: 10.1016/j.techfore.2012.04.013

  • CrossRef
  • Google Scholar

• 97

  WhiddettS.HollyfordeS. (2003). A Practical Guide to Competencies: How to Enhance Individual and Organisational Performance. Chartered Institute of Personal and Development. Available online at: https://books.google.co.za/books?hl=enandlr=andid=ZP850cgpGI0Candoi=fndandpg=PR9anddq=Whiddett,+S.,+%26+Hollyforde,+S.+(2003).+A+Practical+Guide+to+Competencies:+How+to+enhance+individual+and+organisational+performance.+London:+CIPD.andots=Me25XyMacPandsig=pBzsEhDebA0y (Accessed February 5, 2025).

  • Google Scholar

• 98

  WhiteR. W. (1959). Motivation reconsidered: the concept of competence. Psychol. Rev.66, 297–333. doi: 10.1037/h0040934

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 99

  WillcocksL. P. (2024). Automation, digitalization and the future of work: a critical review. J. Electron. Bus. Digit. Econ.3, 184–199. doi: 10.1108/JEBDE-09-2023-0018

  • CrossRef
  • Google Scholar

• 100

  World Economic Forum (2025). Future-of-jobs-report-2025. World Economic Forum. Available online at: https://www.weforum.org/reports/the-future-of-jobs-report-2025/ (Accessed January 30, 2025).

  • Google Scholar

• 101

  WyrzykowskiR. (2023). Rural-urban digital divide still poses a challenge in South Africa. Opensignal. Available online at: https://insights.opensignal.com/2023/12/13/rural-urban-digital-divide-still-poses-a-challenge-in-south-africa (Accessed April 26, 2026).

  • Google Scholar

• 102

  YaoY.CaoJ.TanZ.ChenK. (2025). Developing a digital health competency assessment framework for public health services: a Delphi-AHP approach. Front. Public Health13, 1–11. doi: 10.3389/fpubh.2025.1634261

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 103

  ZenkL.PausitsA.BrennerB.CampbellD. F. J.BehrensD. A.StöcklerE. M.et al. (2024). Meta-competences in complex environments: an interdisciplinary perspective. Think. Ski. Creat.53, 1–13. doi: 10.1016/j.tsc.2024.101515

  • CrossRef
  • Google Scholar

• 104

  ZhangW.TaoX. Bin, Fan, X. L.WangA. P. (2025). Development of evaluation index system for functional ability of older patients with stroke based on healthy aging: a modified Delphi study. Front. Public Health13, 1–11. doi: 10.3389/fpubh.2025.1562429

  • Pubmed Abstract
  • CrossRef
  • Google Scholar