Original Research

Blended learning framework for management competencies in South African retail

Peter A. Champion

Received: 12 Apr. 2026; Accepted: 02 June 2026; Published: 14 July 2026

Copyright: © 2026. The Author. Licensee: AOSIS.
This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license (https://creativecommons.org/licenses/by/4.0/).

Abstract

Orientation: The W&R SETA identified retail manager and retail supervisor as the most difficult-to-fill vacancies in South Africa, citing deficits in experience, qualifications and skills. No fit-for-purpose blended learning framework for this context existed.

Research purpose: This study aimed to develop, implement and evaluate the effectiveness of a blended learning framework for developing critical management competencies in a large South African hardware retailer.

Motivation for the study: Existing models did not account for South Africa’s limited Information Technology (IT) infrastructure, low digital literacy, geographic dispersion or operational constraints preventing extended off-site training.

Research approach/design and method: A quantitative pre–post survey design was used with a purposive sample of 94 management development programme participants across six South African regions. A 50-item questionnaire measuring nine competency constructs was completed by learners and line managers before training and 6 weeks post-training. Scale reliability, factor analysis and Analysis of Variance (ANOVA) established reliability and validity.

Main findings: Improvements were recorded for all nine constructs – communication, motivation, conflict management, problem-solving, time management, decision-making, goal-setting, delegation and self-awareness – with a mean improvement of 11.2% (learner) and 12.3% (line-manager). All 94 participants were assessed as competent against the South African Qualifications Authority (SAQA) National Qualifications Framework (NQF) Level 4 unit standards by accredited Wholesale & Retail Sector Education and Training Authority (W&R SETA) assessors.

Practical/managerial implications: Seven framework refinements – on-the-job primacy, role differentiation, structured coaching, pre-reading orientation, social interaction, e-learning as a complement and distributed assessment – provide actionable design principles for training practitioners in resource-constrained retail environments.

Contribution/value-add: This study produced an empirically evaluated blended learning framework, the first of its kind reported in the literature for the South African retail context, grounded in adult learning theory and responsive to W&R SETA management competency priorities.

Keywords: blended learning; management competencies; South African retail; adult learning; andragogy; experiential learning; W&R SETA; on-the-job training.

Introduction

Orientation

South Africa’s wholesale and retail sector is one of the country’s principal sources of formal employment. At the time this study was conducted, the sector employed approximately 3.2 million workers, constituting 21% of the active labour force, across approximately 87 790 enterprises ranging from single-outlet micro-retailers to national chains with hundreds of branches (Wholesale and Retail Sector Education and Training Authority [W&RSETA], 2019). Despite this scale, the sector confronted a persistent and well-documented shortage of competent frontline managers. The 2019–2020 W&R SETA Sector Skills Plan identified retail manager (general) and retail supervisor as among the most difficult vacancies to fill, attributing the difficulty to three interrelated deficits: lack of experience (28%), lack of qualifications (28%), and lack of skills (27%) (W&R SETA, 2019). The same plan highlighted critical management skills requirements, including decision-making, interpersonal skills, emotional intelligence, assertiveness, teamwork, time management, planning, conflict management, analytical skills and problem-solving.

A second structural challenge compounded the training problem. Both Small, medium and Micro Enterprises (SMME) and large-format retail outlets were typically staffed with a minimum operational headcount. Releasing employees for extended off-site classroom training was operationally untenable for the majority of retailers, regardless of their size. The W&R SETA’s (2021) own evaluation of training interventions found that 89% of retail workplaces had not been capacitated to understand or implement blended learning, and called explicitly for a new training delivery model that would allow skills development to proceed without requiring employees to leave the workplace for extended periods.

Technology-mediated learning appeared to offer a partial solution. The W&R SETA’s (2017) project report noted that international evidence pointed strongly towards blended approaches combining e-learning with face-to-face and workplace learning as the dominant model in comparable retail environments. UK surveys indicated that 84% of retailers employed a blended approach, while Canadian figures reached 96%. However, South African uptake remained limited: only 34% of large retailers and 33% of SMMEs were using e-learning as part of their training mix (W&R SETA, 2017). The principal barriers were inadequate IT infrastructure (present in only 41% of large retailers), insufficient employee digital literacy, and the complete absence of a formal sector blended learning policy.

This convergence of sector-specific competency needs, operational constraints on training delivery, partial technology readiness and the absence of contextually appropriate frameworks constitutes a clearly delineated research gap: no empirically validated, fit-for-purpose blended learning framework has been developed for the South African retail management context. The existing literature on blended learning and management competency development, while extensive internationally, contains no study that addresses the simultaneous constraints of limited IT infrastructure, low digital literacy, geographic dispersion and operational restrictions on off-site training within an African retail environment. This study directly addresses that gap. The researcher, drawing on extensive experience as a senior Human Resources (HR) and training practitioner in the South African hardware retail sector, proposed a refined blended learning framework to meet the W&R SETA’s articulated need. The selected organisation, a large hardware retail chain with more than 300 outlets and in excess of 6000 employees across South Africa and several other African countries, provided both the implementation site and the sample population for empirical testing of the framework. Hardware retail was selected as the focus sub-sector because it is one of the major retail categories identified by the W&R SETA (2019), employing a nationally distributed workforce with a demographic profile broadly representative of the broader retail sector.

Research purpose and objectives

The primary objective was to develop a fit-for-purpose blended learning framework for the development of management competencies within a large South African retailer. Three secondary objectives supported this aim:

• to identify learning theories applicable to a blended learning framework for developing management competencies within the selected hardware retail chain;
• to identify generic management competencies required within the South African hardware retail sector; and
• to refine the blended learning framework as an approach for developing management competencies within the South African hardware retail industry.

The study hypothesised that a blended learning framework in which all support and learning-circle elements function as an interrelated, mutually dependent system rather than as independent components would be effective in developing the identified management competencies and producing measurable behavioural change in the workplace.

Literature review

Adult learning theories

Effective adult workplace learning is underpinned by several complementary theoretical traditions. Cognitivism, the dominant general learning paradigm in this study, focuses on the internal mental processes of learning – including thinking, memory, recall and problem-solving (Knowles et al., 2015). Bloom et al.’s (1956) taxonomy, organised across cognitive, psychomotor and affective domains, provided the instructional design scaffold for competency development across the framework. Cognitivist instructional design principles, task-centred instruction and cognitive apprenticeship were particularly salient: the former ensures that all training activities are grounded in realistic workplace tasks, while the latter positions expert role modelling as a mechanism through which novice managers acquire complex competency through observation and guided practice (Collins, Brown & Newman, 1989).

Social learning theory (Bandura, 1977) contributed the proposition that learning is fundamentally a social process mediated by observation, imitation and the modelling of expert behaviour. This theoretical lens directly informed the framework’s coaching and in-store facilitation elements. Humanism, associated with Maslow’s (1943) hierarchy of needs, emphasised the importance of creating a supportive, safe learning environment in which learners can achieve higher-order development, a principle operationalised through the framework’s emphasis on facilitator relationship-building and learner psychological safety.

For adult learners specifically, three theories exerted the greatest influence on framework design. Knowles’s (1984) andragogy proposes six foundational assumptions about adult learners: they need to know the purpose of learning before engaging with it; they are self-directed in their learning orientation; they bring significant prior experience as a learning resource; their readiness to learn is linked to their current social and occupational roles; they are problem-centred rather than subject-centred; and they are most powerfully motivated by internal rather than external factors consistent with self-determination theory (Deci & Ryan, 2000), which identifies autonomy, competence and relatedness as the primary drivers of intrinsic motivation. Each of these assumptions shaped specific framework components: pre-reading satisfies the purpose-clarity requirement; portfolio-based assessment exploits prior experience; and on-the-job application addresses the problem-centred orientation.

Hase and Kenyon’s (2000) heutagogy extends andragogy by arguing that highly capable adult learners are self-determined; they not only direct their own learning but also actively define the competencies they need to develop. The framework accommodated this through structured self-reflection activities and individual competency goal-setting as part of the assessment process. Kolb’s (1984) experiential learning cycle, concrete experience, reflective observation, abstract conceptualisation and active experimentation, provided the procedural sequence for the learning circle component of the framework. On-the-job training provided the concrete experience; discussion boards and facilitator debriefs enabled reflective observation; classroom sessions introduced abstract frameworks; and assessments in the live retail environment constituted active experimentation.

Learner motivation received explicit theoretical attention. Driscoll (2005), supported by Bates (2015), proposed four motivational conditions: the learner’s attention must be captured; the relevance of content must be apparent; the learner must develop confidence in their ability to perform; and the learner must derive a sense of satisfaction from outcomes. These conditions were built into the framework through structured pre-reading (relevance), structured facilitation (attention and confidence), and recognised portfolio attainment (satisfaction). Merrill (2013) confirmed that motivation is not peripheral but constitutive of learning success, a finding that elevated motivational design to the same priority level as content design in this study.

Management competencies in South African retail

The identification of the competencies to be developed was guided by two complementary sources: the W&R SETA’s sector-level skills needs analysis and a review of established leadership development models. The W&R SETA (2019) Sector Skills Plan explicitly identified decision-making, interpersonal skills, emotional intelligence, assertiveness, teamwork, time management, planning, conflict management, analytical skills and problem-solving as scarce and critical competencies at the retail management level.

These sector-level needs were cross-referenced with four leadership development models to ensure theoretical grounding. The Emotionally Intelligent Leadership Development Model (Shankman et al., 2015) emphasises consciousness of self, others and context, directly mapping to the self-awareness, communication and motivation competencies in this study. The Social Change Leadership Development Model (Komives & Wagner, 2017) foregrounds collaboration, common purpose, and citizenship reflected in conflict management and goal-setting competencies. The Complexity Leadership Development Model (Flinn, 2018) addresses leadership co-created through social interaction, informing the framework’s high-density social interaction design principle. The Full-Range Leadership Development Model (Avolio, 2011) emphasises situational awareness and the balance between transformational and transactional leadership, informing the decision-making and delegation competencies.

This triangulation yielded nine management competencies operationalised against SAQA NQF Level 4 unit standards: communication, motivation, conflict management, problem-solving, time management, decision-making, setting goals and objectives, delegation and self-awareness. Exploratory factor analysis subsequently confirmed that these nine competencies organised into two higher-order factors: Doing (interpersonal and cognitive-reactive competencies) and Planning (task-organising and execution-directing competencies), consistent with established distinctions in management competency literature.

Blended learning: Models and the South African context

A review of the principal blended learning models in use internationally informed both the framework design and the identification of required refinements. The Self-Directed Model allows learners to progress through online content at their own pace, with classroom contact used to consolidate and extend learning. The Supplemental Model retains traditional classroom instruction as the primary modality, with online content serving an enrichment function. The Mastery-Based Model gates progression to the next module on demonstrated competency attainment in the current one. The Enriched Virtual Model reverses the traditional ratio, with the majority of learning delivered online and face-to-face sessions used for activities that benefit from synchronous interaction (Botha & Coetzee, 2013; Launer, 2010).

All four models share a significant limitation for the South African retail context: they place technology at the centre of the learning architecture, creating an inherent dependency on reliable IT infrastructure, learner digital literacy and adequate internet connectivity, all identified as problematic within the South African retail sector by the W&R SETA (2017). A second shared limitation was the relatively minor role accorded to on-the-job training and coaching, despite well-established evidence that up to 70% of effective workplace learning occurs through on-the-job experience and mentoring interactions (Botha & Coetzee, 2013; Noe, 2017). A third limitation was the inadequate incorporation of adult learning principles, particularly andragogical requirements for social interaction, purpose-clarity, and experiential relevance and the neglect of coaching and mentoring as modelling mechanisms (Dias et al., 2014; Wang, 2018). These three limitations collectively defined the design space for the seven refinements introduced in the present framework.

The framework proposed in this study comprises two interdependent main components: a Support Circle and a Learning Circle. The Support Circle encompasses three enabling roles: the facilitator (who simultaneously serves as subject matter expert, relationship builder, coach and standard setter), the e-learning and online collaboration platform (delivering pre-reading and hosting asynchronous discussion), and a dedicated workplace coach or mentor. The Learning Circle operationalises Kolb’s (1984) experiential cycle through five sequential phases: pre-reading (activating prior knowledge and establishing learning purpose), classroom or virtual facilitation (structured theoretical input and peer interaction), on-the-job training (majority of learning hours, in situ with facilitator and coach support), planned in-store facilitation (structured observation and modelling in the live retail environment) and assessment (formative and summative, distributed over 6 weeks). The two components are deliberately interdependent: the learning circle cannot function effectively without functioning support structures, and the support structures have no value if the learning circle phases are not all implemented. The facilitator is the keystone element connecting both components.

Research methods and design

Research approach

The study adopted a positivist epistemology and realist ontology, consistent with its objective of measuring observable behavioural change through quantifiable pre–post differences. A quantitative research strategy was employed, using a survey approach as described by Saunders et al. (2016). This paradigmatic positioning was congruent with the researcher’s view that management competency application constitutes a measurable social reality that exists independently of individual perceptions of it. The philosophical stance of realism holds that organisations are social entities with structures and processes that create a shared, if socially influenced, reality, one that can be studied using the scientific methods of positivism (Saunders et al., 2016). The choice of a quantitative, survey-based design was also pragmatically motivated: it enabled the collection of standardised, comparable data from a geographically dispersed sample across six regions simultaneously, and it produced numerical outcomes that could be subjected to the inferential statistical procedures required to establish reliability, validity, and the significance of pre–post change.

Research method

Research participants

The study was conducted within a single large South African hardware retail chain with more than 300 outlets and in excess of 6000 employees across South Africa and several African countries. The target population comprised branch managers, assistant managers, and employees identified as having future management potential who had been enrolled on the organisation’s management development programme (MDP), totalling 154 individuals. Coronavirus disease 2019 (COVID-19) lockdown restrictions reduced the final sample to 94 participants who had completed training prior to the implementation of social-distancing measures. This sample size was adequate given Comrey and Lee’s (1992) criterion of a minimum of 10 observations per measured construct across the study’s nine constructs. Table 1 presents the demographic profile.

Measuring instruments

The researcher developed a 50-item structured self-administered questionnaire organised into nine competency scales. Items were adapted from previously validated scales and cross-referenced against the learning outcomes of six SAQA NQF Level 4 unit standards: unit standard 14667 (describe and apply management functions), 242821 (team leader responsibilities), 242819 (motivate and build a team), 242811 (prioritise time and work), 114589 (manage time productively) and 242824 (apply leadership concepts in the workplace). Parallel learner self-rating and line-manager observer-rating forms were used, providing two independent perspectives on the same competency application. Items were scored on a five-point Likert-type scale ranging from 1 (not effective or not able) to 5 (highly effective or always able). Each competency scale comprised between five and six items, with total scores calculated as scale means to allow comparability across constructs of differing item counts. The adaptation process drew on previously published instruments assessing managerial effectiveness (Noe, 2017) and leadership competency (Avolio, 2011), with items reworded to reflect NQF Level 4 behavioural descriptors and retail-specific operational contexts. Content validity was established through review by three subject matter experts: a senior W&R SETA skills development facilitator, a registered assessor, and an HR academic who independently rated each item for relevance and clarity against the unit standard learning outcomes. Items with disagreement ratings were revised before finalisation. The pilot phase, conducted with a separate cohort of 12 management trainees not included in the main study, produced no items flagged for ambiguity.

The questionnaire was preceded by information-sharing sessions with participants and their line managers to clarify item meaning, explain the purpose of the research, and mitigate potential response bias. Participants were explicitly informed that completion was voluntary, responses would be treated anonymously, and there would be no career consequences attached to any particular response pattern (Dillman et al., 2014).

Research procedure

The instrument was administered at two time points: immediately prior to the training intervention (pre-training) and 6 weeks after the completion of the final training module (post-training). The 6-week interval was deliberately selected to allow sufficient time for learners to attempt to apply newly acquired skills in the workplace, providing ecologically valid evidence of competency transfer rather than immediate post-training recall.

The training intervention comprised three W&R SETA-accredited NQF Level 4 skills programmes delivered across a 3-month period using the proposed blended learning framework. Independent criterion-referenced validation was provided through portfolios of evidence submitted by each participant, assessed by W&R SETA-accredited assessors; 10% of all portfolios were further moderated by an independent W&R SETA-accredited moderator.

Ethical considerations

Ethical clearance was obtained from The Da Vinci Institute for Technology Management Institutional Review Board prior to data collection. All procedures were conducted in accordance with the 1964 Helsinki Declaration and its later amendments. Written informed consent was obtained from all individual participants included in the study.

Statistical analysis

Cronbach’s alpha (α) and McDonald’s omega (ω) assessed scale reliability for all nine constructs across both time points and both rater groups. Pearson product-moment correlations established inter-item reliability and inter-scale associations. One-way ANOVA examined whether competency outcomes differed systematically by facilitator, geographic region or gender. Differential Item Functioning (DIF) analysis at the item level confirmed that no individual questionnaire item systematically advantaged or disadvantaged any demographic sub-group, supporting measurement fairness. Exploratory Factor Analysis (EFA) with principal axis factoring was applied to post-training data to examine the underlying factor structure of the nine competency constructs and assess construct validity (Thompson, 2004).

Results

Reliability and validity of the measuring instrument

Cronbach’s alpha and omega coefficients exceeded 0.70 for all nine competency scales across both time points and both rater forms, confirming adequate internal consistency throughout. Alpha values ranged from α = 0.71 to α = 0.89 across scales and time points, with the majority of scales achieving α ≥ 0.78, indicating satisfactory to good internal consistency (Nunnally, 1978). McDonald’s omega coefficients were comparable, ranging from ω = 0.73 to ω = 0.91, confirming that alpha values were not inflated by item redundancy. Pearson correlations between items within each scale were positive and statistically significant, confirming that all items reliably measured their intended constructs. The overall Pearson correlation analysis indicated good reliability for the full instrument.

Face and criterion validity were confirmed through expert review by subject matter specialists and cross-referencing of item content against SAQA unit standard learning outcomes, respectively. Content validity was confirmed by the close alignment between required workplace behaviours specified in the SAQA unit standards and the behavioural indicators assessed in the questionnaire items. ANOVA results yielded no statistically significant competency outcome differences by geographic region or gender, indicating that the framework was equitable in its effects across the diverse demographic and geographic composition of the sample. Facilitator experience level did account for meaningful variance in outcomes, with facilitators holding more than 2 years of adult-facilitation experience associated with higher gains a finding that reinforced the theoretical centrality of the facilitator role identified in the literature.

DIF analysis confirmed measurement fairness at the item level: no individual item systematically disadvantaged any racial or gender sub-group. EFA identified two stable factors. Factor 1 (Doing) comprised six constructs: problem-solving, communication, decision-making, motivation, conflict management and self-awareness. Factor 2 (Planning) comprised three constructs: delegation, setting goals and objectives and time management. The Keiser-Meyer-Olkin (Measure of sampling adequacy) (KMO) measure of sampling adequacy was satisfactory (KMO = 0.83), and Bartlett’s test of sphericity was statistically significant (p < 0.001), justifying the application of factor analysis. The two-factor solution accounted for 61.4% of the total variance in post-training competency scores. All factor loadings exceeded 0.45, and no item demonstrated cross-loadings above 0.30, confirming discriminant validity between the two factors. High inter-competency correlations across all nine constructs supported the theoretical position that management competencies function as an interrelated system and should be trained as a cluster rather than in isolation.

Competency improvement

Table 2 presents the mean percentage improvement across all nine competency constructs as rated by learners and line managers from pre- to post-training. Positive improvement was recorded for every construct by both rater groups.

Precise values per questionnaire item are reported in Champion (2021).

Conflict management registered the largest overall gain (14.4% for both rater groups). Communication produced the lowest line-manager-assessed gain (9.6%), while self-awareness was the lowest learner self-assessed gain (9.4%). In all constructs, learners rated their own improvement more generously than line managers a pattern attributable to self-enhancement bias and the inherent delay between skill acquisition and observable behavioural change in the workplace (Dillman et al., 2014). Critically, both rater groups agreed on the direction of improvement for every construct, providing convergent validity for the overall finding. The overall mean improvement of 11.2% (learner) and 12.3% (line manager) is practically meaningful given that these were novice managers developing foundational competencies through a 3-month programme in a live operational setting.

All 94 participants (100%) were assessed as competent against all applicable SAQA NQF Level 4 unit standards by accredited W&R SETA assessors, with an independent moderation sample confirming consistency in assessment decisions. This result warrants contextualisation: criterion-referenced competency assessment in SAQA-accredited programmes sets a defined minimum performance threshold per unit standard, rather than ranking learners against each other. The 100% attainment rate reflects that all participants met or exceeded that minimum threshold, not that all achieved equal or perfect performance. Portfolio assessment decisions were based on evidence of workplace application across multiple competency demonstrations over 6 weeks, assessed by external accredited assessors who had no involvement in delivering the training. The independence of the assessment function from the instructional function is a design feature of the seventh framework refinement and is critical to interpreting this result as externally validated evidence rather than trainer-generated self-confirmation. This criterion-referenced attainment rate provides external, independently verified validation of the framework’s effectiveness, complementing and strengthening the questionnaire-based evidence. The portfolio assessment process required learners to demonstrate practical application of each competency in their actual workplace setting, not merely reproduce theoretical knowledge under controlled conditions. This assessment design, embedded within the seventh framework refinement, ensured that competency attainment reflected genuine behavioural change in a live retail management context rather than training-room performance, a distinction of considerable practical significance for the W&R SETA’s workforce development mandate.

Discussion

Outline of results

The proposed blended learning framework was effective in developing all nine identified management competencies within a large South African hardware retailer, confirming the study’s primary objective and supporting the research hypothesis. Both within the quantitative pre–post measurements and the externally adjudicated SAQA competency attainment data, the evidence consistently pointed to meaningful competency development. A mean overall improvement of 11.2% (learner self-rated) and 12.3% (line-manager-rated) represents a practically significant outcome: these gains were achieved by frontline managers with limited prior formal management training, working in live retail environments, over a 3-month programme, under conditions that are considerably more demanding than controlled training settings. The 21 literature findings, seven framework refinements and 11 empirical findings derived during the study collectively explain how and why the framework achieved these results.

The seven refinements merit detailed consideration, as they represent the substantive contribution of this framework over existing models. The first refinement, deliberate high-density social interaction, addressed the motivational deficits characteristic of asynchronous e-learning (Dias et al., 2014) by building structured peer-to-peer and learner–facilitator interaction into every phase of the learning circle. This directly instantiated Bandura’s (1977) social learning mechanisms and Knowles’s (1984) andragogical requirement for a participatory learning environment. The second refinement, on-the-job primacy, ensured the majority of learning hours were spent in the workplace, with facilitators and coaches travelling to the learner’s site. This not only eliminated the operational barrier of extended time away from work but maximised physical and psychological fidelity between the training environment and the performance environment, directly operationalising Kolb’s (1984) concrete experience phase.

The third refinement, e-learning as a complement, positioned technology as an enabler of pre-reading and asynchronous discussion rather than as the primary delivery vehicle. This design was not only appropriate given South Africa’s infrastructure constraints but theoretically sound: the literature review confirmed that e-learning approaches often fail precisely because they place technology at the centre rather than the learner (Wang, 2018). The fourth refinement, role differentiation, distributed subject matter expertise, facilitation and coaching across distinct role-players, preventing the single-facilitator overload that had been identified as a failure mode in existing models. The fifth refinement, pre-deployment role-player training, ensured that all facilitators and coaches were prepared in adult learning principles before any learner-facing contact, consistent with the empirical finding that facilitator experience level was the principal source of variance in learner outcomes. The sixth refinement, pre-reading orientation, satisfied the andragogical purpose-clarity requirement (Knowles, 1984) and improved the quality of classroom engagement by activating relevant prior knowledge. The seventh refinement, a distributed, workplace-based assessment over 6 weeks, operationalised the principles of practice and overlearning, spaced consolidation and ecologically valid evidence collection.

The two-factor competency structure (Doing versus Planning) identified through EFA provides a theoretically coherent and practically useful taxonomy for curriculum design and individual development planning. Practitioners may sequence training by addressing foundational Doing competencies, which are more visible and immediately applicable in daily managerial interactions, before introducing Planning competencies, which require sustained practice over a longer period to embed.

Practical implications

Several practitioner-level insights emerged from the researcher’s systematic reflective observations during framework implementation. Senior management endorsement was non-negotiable as a precondition for framework success: without visible and active support from senior leadership, learners were unlikely to be granted the time, latitude and psychological safety required to practise newly acquired competencies in the workplace. This finding echoes broader evidence that organisational climate is a critical transfer variable in training effectiveness (Noe, 2017).

The calibre and developmental orientation of appointed coaches and mentors exerted a decisive influence on on-the-job learning quality. Coaches required not only subject matter competence but an orientation towards developing people, a finding with direct implications for selection criteria. The researcher’s observation that at least 2 years of adult-facilitation experience was required for independent facilitator deployment implies that organisations should build a structured facilitator development pipeline rather than treating facilitator deployment as an ad hoc arrangement.

The finding that all role-players, facilitators, coaches, subject matter experts and learners required a practice-grounded understanding of competency attainment, not merely theoretical familiarity, points to an important design principle for future implementations: orientation activities for all role-players should include workplace demonstrations or simulations of what each competency looks like in action at the required performance level.

The W&R SETA (2021) recommended the development of a formal sector blended learning policy. The present framework provides an empirically tested foundation for such a policy. However, the researcher’s observation that the working retail environment is inherently fluid implies that any formal policy must preserve operational flexibility and avoid imposing bureaucratic uniformity that would undermine contextual adaptation.

Limitations and recommendations

Four principal limitations bound the interpretation of these findings. Firstly, the sample was drawn from a single hardware retail chain; while the framework’s design principles are theoretically generalisable, the extent to which findings transfer to other retail sub-sectors, such as food, clothing, pharmaceutical and household goods retail, or to non-retail vocational training contexts, requires empirical verification. The seven framework refinements are grounded in adult learning theory, which has demonstrated cross-sectoral applicability (Knowles et al., 2015); however, sector-specific factors such as shift-work patterns, unionisation levels, franchise governance structures and digital infrastructure availability may modify both the implementability and the outcomes of the framework in other contexts. Practitioners intending to apply the framework beyond hardware retail are advised to conduct a formal context analysis, including technology readiness, facilitator availability and operational scheduling constraints before implementation. Secondly, the absence of a randomised control group meant that alternative explanations for competency gains, including maturation, history effects and Hawthorne effects, could not be fully excluded. The single-group pre–post design is a recognised limitation of applied workplace research, where random assignment to a no-treatment control condition is frequently precluded by organisational and ethical constraints (Saunders et al., 2016). The convergence of two independent data sources, quantitative self- and manager-report questionnaires and externally assessed SAQA portfolios, provides stronger inferential support for the framework’s effectiveness than either source alone, but does not fully substitute for experimental controls. Thirdly, COVID-19 reduced the target sample by 39%, and while 94 participants were statistically sufficient, the possibility of selection bias in the reduced sample cannot be dismissed. Fourthly, both self-report and manager-report measures are susceptible to response bias, although the independent SAQA portfolio assessment rate of 100% competency attainment provided a partial check on this limitation.

Future research should: test the framework using a randomised controlled design with an active comparison condition; extend the framework to other retail sub-sectors and to non-retail South African vocational training contexts; incorporate longitudinal follow-up beyond 6 weeks to examine competency sustainability; triangulate self- and manager-report data with objective performance metrics such as skills audit findings and appraisal ratings; and explore how emerging mobile learning and AI-assisted coaching tools may be incorporated into the framework without displacing the human-centred, on-the-job orientation identified here as its critical differentiating strength.

Conclusion

This study demonstrated that a blended learning framework placing the adult learner, on-the-job practice, and facilitated social interaction at its centre, while using technology as an enabling rather than a driving force, produced significant and independently verified improvements in all nine critical management competencies identified for the South African retail sector. The framework directly addressed the W&R SETA’s (2019, 2021) articulated need for a fit-for-purpose alternative training delivery model that does not require extended workplace absence. Its seven evidence-based refinements constitute replicable design principles for future blended learning development, applicable both within South African retail and across the broader domestic vocational training landscape. The convergent evidence from pre–post questionnaire data and externally adjudicated SAQA competency portfolios provides a credible empirical foundation for this conclusion, while the study’s acknowledged limitations, particularly the single-organisation sample and the absence of an experimental control, frame the appropriate boundaries of generalisation. Future research employing randomised or quasi-experimental designs across multiple retail sub-sectors will be necessary to establish the full transferability of these findings. For the W&R SETA, the framework offers an evidence-informed starting point for sector blended learning policy development.

Acknowledgements

During the preparation of this work, the author used Claude (Anthropic, claude-sonnet-4–6) to support language generation and text drafting. The content was reviewed and edited by the author, who takes full responsibility for its accuracy. All conceptual development, theoretical framing, construct definitions, proposition development, interpretation of literature and final scholarly judgements were the sole responsibility of the author.

This article is based on research originally conducted as part of Peter A. Champion’s doctoral thesis titled, ‘A fit for purpose blended learning framework for the development of management competencies in a large South African retailer’, submitted to the Faculty of Business Management of The Da Vinci Institute in 2021. The thesis was supervised by Professor HB Klopper. The thesis was reworked, revised and adapted into a journal article for publication. The author confirms that the content has not been previously published or disseminated and complies with ethical standards for original publication.

The author, Peter A. Champion, declares that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

Peter A. Champion: Conceptualisation, Formal analysis, Methodology, Writing- original draft, Writing- review and editing. The author confirms that this work is entirely their own, has reviewed the article, approved the final version for submission and publication, and takes full responsibility for the integrity of its findings.

The author received no financial support for the research, authorship and/or publication of this article.

Data sharing is not applicable to this article as no new data were created or analysed in this study.

The views and opinions expressed in this article are those of the author and are the product of professional research. They do not necessarily reflect the official policy or position of any affiliated institution, funder, agency or that of the publisher. The author is responsible for this article’s results, findings, and content.

References