Work-Integrated Learning (WIL) is increasingly recognized not merely as a pedagogical approach but as a transdisciplinary field that explores the dynamic relationship between work and learning across institutional and societal boundaries (Areskoug Josefsson et al., 2024; Lundh Snis et al., 2025). Emerging literature positions WIL at the forefront of addressing global societal challenges, fostering collaborative knowledge generation among Higher Education Institutions (HEIs), students, professionals, and external stakeholders (Lundh Snis et al., 2025; Sunnemark et al., 2024). In today’s rapidly evolving technological landscape, shifting labour market demands, and growing expectations for sustainable and socially responsible innovation in higher education, WIL must emphasize practical relevance and conceptual integrity to shape the future trajectory of higher education (Vallo Hult et al., 2025). Lundh Snis et al. (2025) argue that WIL should be viewed as a collaborative, ecosystemic process that democratizes knowledge production and engages diverse actors in co-creating solutions to complex societal issues. Furthermore, WIL has been framed not only as a seamless integration of academia and professional life (Bernhard & Olsson, 2023) but also as a participatory model that enhances broader societal engagement and contributes to sustainable development (Lundh Snis et al., 2025; Gankhuyag et al., 2025). This dual perspective extends the understanding of WIL beyond a narrow focus on employability, viewing it instead as a relational and governance-intensive ecosystem in which universities, organizations, and communities collaboratively produce knowledge and value. This broadened perspective necessitates further theoretical exploration and robust empirical assessment of the relational processes that sustain effective collaboration across institutional divides (Vallo Hult et al., 2025).

As HEIs undergo significant transformations, including digital innovation, the integration of artificial intelligence, cybersecurity commitments, internationalization pathways, and accountability for graduate employability (Amanzhol et al., 2024; Msomphora, 2025; Parambil et al., 2024; Schueller, 2023), there is an urgent need for models that bridge theoretical understanding with practical application while maintaining academic integrity and societal responsiveness (Jackson & Dean, 2023). In this context, WIL is ideally positioned as a promising link between theory and practice. However, critical perspectives suggest that its conceptualizations may be too diverse and occasionally lack sufficient theoretical grounding, advocating for a multidimensional understanding of WIL (Björck & Willermark, 2025; Vallo Hult et al., 2025). Consequently, several authors have endeavoured to define perspectives on WIL more rigorously. In their work, Sunnemark et al. (2024) argue that WIL should not be viewed in isolation from placements, but rather as an organized mode of collaborative knowledge-building that requires careful alignment of pedagogical design with stakeholder interaction and institutional governance. This perspective echoes calls for more consistent conceptual frameworks to ensure the long-term quality of WIL as perceived by stakeholders (Dean et al., 2020; Dollinger & Brown, 2019; Vallo Hult et al., 2025). Without such frameworks, WIL risks fragmentation and heterogeneity, diluting its core concepts.

Framing WIL in this manner means considering a governance-intensive positioning understood not as an imposition of bureaucratic control, but as structured and relational coordination mechanisms essential for fostering principled engagement and shared responsibility among institutions. Within this evolving landscape, understanding how value is co-constructed in WIL ecosystems becomes critical. Service-Dominant Logic posits that value is generated through the integration of resources among multiple stakeholders, rather than being delivered unilaterally (Vargo & Lusch, 2004). This relational perspective aligns well with sociocultural and practice-based learning theories, which view knowledge as derived from participation in social actions and shared practices. Wenger-Trayner and Wenger-Trayner (2020) specifically frame value creation as occurring within social learning environments where members collectively negotiate meaning, align practices, and generate value through collaborative engagement. In WIL settings, universities, students, companies, and facilitators contribute collaborative resources – such as expertise, networks, problem framings, and organizational capabilities – that must be integrated into existing social learning contexts to yield meaningful outcomes. Recent conceptual synthesizations of co-creation in higher education underscore the emergent nature of value in interactive, dialogic, and agency-focused processes (Omland et al., 2025).

From another standpoint, Prahalad and Ramaswamy (2004) operationalized co-creation through the DART paradigm, which emphasizes Dialogue, Access, Risk Assessment, and Transparency as fundamental elements of the co-creation process. Developed it within the field of marketing and innovation management and grounded in the principles of Service-Dominant Logic, the model reflects a shift from traditional, firm-centric views of value creation, where value is embedded in products and delivered to passive consumers, towards a relational perspective in which value emerges through interactions among multiple actors (Nagarethenam et al., 2018). Taghizadeh et al. (2016) then developed a scale to assess these dimensions in business innovation contexts. The model remains an evolving research domain as it is adopted to diverse areas of application, due to its conceptual clarity and flexibility, allowing for multiple configurations depending on organizational and interactional settings (Hurtado-Cardona et al., 2023). However, empirical support for the DART framework remains limited, and applications in higher education WIL ecosystems are scarce. This gap is significant, as university-industry collaborations differ structurally from commercial innovation scenarios, particularly in terms of pedagogical mediation, assessment regimes, and shared accountability for learning outcomes (Mackaway & Winchester-Seeto, 2018; Vallo Hult et al., 2025). On the other hand, recent WIL studies have identified time, trust, and territory as “facilitating factors” that encourage collaborative learning across organizational boundaries (Gankhuyag et al., 2025). Trust, especially in cross-boundary collaborations, is often characterized as the willingness to embrace vulnerability in interdependent relationships, which correlates with improved collaboration and communication in university-society engagement studies (Law & Le, 2023). These relational dynamics align closely with the concept of psychological safety – the extent to which participants feel free to express ideas, questions, and concerns without fear of relational harm (Edmondson, 2018). In increasingly multidisciplinary and multicultural WIL ecosystems, fostering psychological safety is crucial for maintaining meaningful communication and enabling collaborative experimentation. Moreover, WIL ecosystems function within larger stakeholder governance frameworks. Stakeholder theory conceptualizes organizations as embedded within networks of relationships that require balanced value creation among various actors (Freeman et al., 2020). In education, this implies that the quality of WIL should not be evaluated solely on student outcomes but also on how external stakeholders, institutions, and society perceive its value. Governance perspectives emphasize coordination mechanisms that enable distributed actors to align objectives, share responsibility, and manage interdependencies (Jacobides et al., 2018).

From these theoretical vantage points, tools that can account for relational quality from stakeholder perspectives are necessary to assess the effectiveness of WIL. However, quantifying such relational constructs poses significant methodological challenges. Measurement theory highlights the importance of establishing construct validity and measurement invariance, as dimensional structures may become unstable if transferred from one context to another (Henseler et al., 2016). Instruments not designed for a specific environment may need empirical re-evaluation before being applied in different relational contexts. Tensions in scale validation are evident in the debate between empirically grounded and empirically derived approaches, such as Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA). While CFA assumes a predefined dimensional structure, EFA allows latent dimensions to emerge inductively from context-specific data, especially in newly established empirical settings (Howard, 2016). If the DART framework is applied to WIL, it remains theoretically uncertain whether the original four dimensions will retain their distinctiveness or adapt to the relational dynamics of higher education ecosystems.

This study aims to address these conceptual and methodological gaps by translating and empirically assessing a Portuguese adaptation of the DART scale within the WE: Working and Envisioning Community of Practice at Bragança Polytechnic University (Instituto Politécnico de Bragança, IPB). The WE program employs Complex Challenge-Based Learning approaches, in which interdisciplinary and multicultural student teams collaborate with community-based stakeholders for 8-12 weeks to develop solutions to real-world issues. This structured yet relationally rich environment provides a relevant empirical context for exploring how external partners perceive co-creation processes in WIL. By offering empirical insights into the dimensional construction of the DART framework, this study seeks to contribute to the WIL literature in three interconnected ways. First, by investigating whether existing co-creation constructs remain valid in educational contexts. Second, by extending quality establishment by examining how a diagnostic tool can enhance stakeholder engagement in WIL activities. Third, by responding to calls for a stronger theoretical and empirical foundation for WIL as a field capable of addressing complex societal challenges through collaborative knowledge generation (Lundh Snis et al., 2025). Ultimately, this paper positions WIL not merely as a pedagogical strategy but as a governance-driven, stakeholder-oriented ecosystem that requires robust theoretical foundations and empirically based tools to thrive in the future.

Work-Integrated Learning (WIL) is increasingly being recognized as a significant area of research and practice that transcends traditional pedagogical frameworks. It encompasses the dynamic nature of learning that occurs across institutional, professional, and societal boundaries. Current scholarship emphasizes the importance of viewing WIL not merely as a model for placements or a tool for enhancing employability, but as an intricate, governance-driven process through which knowledge, responsibility, and agency are collaboratively constructed among various stakeholders (Sunnemark et al., 2024; Björck & Willermark, 2025; Lundh Snis et al., 2025). This evolving perspective responds to the growing concern over the instrumentalization of WIL, which may reduce it to a mere means of meeting labour market demands or a shortcut to applied competence (Björck & Willermark, 2025). Instead, contemporary literature positions WIL as a relational learning framework where universities, students, organizations, and communities collaboratively engage in problem identification and value creation within complex ecosystems (Gankhuyag et al., 2025; Vallo Hult et al., 2025).

In alignment with the introduction, WIL ecosystems are conceptualized as governance-mediated learning constellations characterized by interdependence and boundary-spanning value production. From this perspective, WIL is not a straightforward transfer of knowledge from academia to practice (or vice versa), but rather the emergence of diverse knowledge configurations resulting from mutual interactions across boundaries. Sunnemark et al. (2024) stress that WIL must be theoretically grounded to avoid “practice-over-theory reductionism,” advocating for a structured approach that connects pedagogical design, stakeholder engagement, and governance mechanisms to ensure conceptual coherence. This insistence on academic rigor is echoed by Björck and Willermark (2025), who identify a common issue of under-theorization within the “WIL” component itself, leading to conceptual dilution.

These critiques underscore the urgency of recognizing WIL as a multidimensional phenomenon in which pedagogical, relational, and institutional dimensions coexist. Viewed through this lens, collaboration is not an adjunct to learning; rather, it serves as its foundational mechanism. Gankhuyag et al. (2025) regard WIL ecosystems as collaborative spaces for social sustainability, where stakeholders negotiate meanings, responsibilities, and expectations over time. In this sense, collaboration within WIL becomes a vital learning tool that enables diverse actors to co-construct their understanding of complex challenges. This aligns with contemporary views of higher education ecosystems, which position universities as integral nodes within innovation networks characterized by interdependence and co-specialization (Jacobides et al., 2018). Value in such ecosystems is not the result of individual contributions but emerges through organized interactions. To view WIL as collaborative learning, it is essential to define the structural conditions that foster generative collaboration rather than mere coordination. Boundary-crossing experiences create these structural conditions. Current WIL research recognizes that learning between academia and the workplace involves negotiating discontinuities in epistemic norms, institutional logics, and professional identities (Sunnemark et al., 2024). Boundary crossing does not imply seamless integration; instead, it introduces productive tensions that turn differences into assets for knowledge creation. However, boundary crossing is not self-sufficient; it relies on mediating artifacts that facilitate interaction and coordination. In WIL contexts, these artifacts may include project briefs, prototypes, assessment rubrics, design documents, and evaluation tools. These artifacts ground dialogue, shape expectations, and transform abstract collaboration into tangible outcomes.

Omland et al. (2025), through their systematic review of co-creation in higher education, illustrate that co-creation occurs interactively through positioning, voice, and negotiated meaning. Artifacts provide the space for such positioning, enabling actors to express interests, challenge interpretations, and redefine obligations. Consequently, boundary objects serve as mediators in the interplay between institutional expectations and emergent learning forms. This perspective allows for a more nuanced understanding of how value is generated within WIL ecosystems. Wenger-Trayner and Wenger-Trayner (2020) theorize value production within social learning spaces, where participants engage in collective meaning-making. Value arises not from isolated contributions but from reasoned engagement in shared practices. In a WIL context, social learning spaces are facilitated by boundary objects that support collaboration while allowing for interpretive flexibility. Rather than relying solely on transactional exchanges, universities, students, and external stakeholders integrate resources and create value through active participation in these learning environments.

Through this framework, the relational dynamics of co-creation can be analytically examined using the DART model, which articulates Dialogue, Access, Risk Assessment, and Transparency as interrelated conditions shaping collaborative value creation (Prahalad & Ramaswamy, 2004). Rather than reiterating its foundational premises, previously asserted in the Introduction, it is important to emphasise that DART functions as a relational structuring device, enabling the analysis of how interaction quality, resource accessibility, shared responsibility, and informational openness collectively configure co-creation processes. Building on service-dominant logic, these dimensions do not operate in isolation; instead, they are enacted through ongoing exchanges among actors who integrate resources within networked environments (Vargo & Lusch, 2004; Nagarethenam et al., 2018). From this perspective, the relevance of DART lies in its capacity to translate abstract notions of co-creation into observable relational conditions. Dialogue supports reciprocal engagement and meaning-making; access structures participation through the availability of resources and information; risk assessment reflects the distribution and negotiation of responsibility; and transparency underpins trust by reducing informational asymmetries. Contemporary literature positions the model as a flexible analytical framework capable of capturing co-creation dynamics across varied organisational settings, while also recognising that its empirical validation remains context-dependent and evolving (Hurtado-Cardona et al., 2023). As such, DART should be understood not as a fixed dimensional structure, but as a conceptual lens whose configuration may shift in response to the relational and governance characteristics of the context in which it is applied.

As such, its application within WIL ecosystems necessitates theoretical refinement. University-industry collaboration involves pedagogical mediation, institutional accountability, and shared ownership of learning outcomes, which differ significantly from commercial innovation contexts. When applied to WIL, the relational dimensions captured by DART may require reconfiguration. In this context, the relationship between compliance and capacity building becomes critical. In governance-intensive WIL ecosystems, particularly those driven by regulatory frameworks, external partners are often entangled in compliance structures, institutional accountability measures, and performance benchmarks. Compliance can thus serve as a primary organizing principle. Nevertheless, contemporary WIL theory posits that true capacity building occurs when organizations internalize compliance structures within social learning processes, rather than merely imposing external compliance (Gankhuyag et al., 2025; Sunnemark et al., 2024). When integrated into boundary-mediated collaboration, compliance mechanisms can foster shared responsibility and clarify expectations, thus promoting capacity building rather than perpetuating a restrictive status quo.

This aligns with Emerson and Nabatchi’s (2015) collaborative governance model, which asserts that principled engagement, shared motivation, and the capacity for joint action are mutually reinforcing elements of successful collaborative systems. Compliance frameworks may stabilize WIL systems while social learning spaces cultivate adaptive capabilities. Recent studies on collaborative performance management indicate that challenges in collaborative governance regimes – such as substantive problem-solving, processual accountability, and goal alignment – are contingent on structured integration among overarching objectives, formal dialogue mechanisms, and evaluative practices. Waardenburg et al. (2025) emphasize that performance management in collaborative settings is not merely a technical issue but also involves interpersonal and iterative processes. Applied to WIL ecosystems, this implies that co-creation instruments must encompass more than transactional exchanges; they should also reflect the quality of relational processes that support learning. In this light, trust emerges as a crucial relational condition. Law and Le’s (2023) systematic review highlights that trust is consistently linked to vulnerability acceptance in interdependent interactions. Trust fosters openness, information sharing, and long-term collaboration. In WIL ecosystems characterized by multidisciplinary and multicultural teams, trust is essential for facilitating authentic dialogue and addressing organizational ambiguities. Trust is closely linked to psychological safety, which is the feeling that it is safe to express ideas and concerns. Psychological safety fosters experimentation and adaptive reflexivity within governance-intensive learning environments.

However, trust alone cannot fully capture the relational prerequisites for co-creation. The concept of territory adds another analytical dimension to WIL theory. Territory can be understood not merely as protective boundaries but as a means of asserting ownership within social learning spaces. According to Wenger-Trayner and Wenger-Trayner (2020), value creation hinges on participants recognizing their legitimate stake in the learning process. As a principle, territory signifies ownership and acknowledges individuals’ value and agency in achieving outcomes. Without this ownership, positioning and voice remain superficial. Empirical studies of immersive, challenge-based WIL initiatives reveal that when participants claim collective artifacts as “our solution” or “our project,” ownership becomes discursively evident, marking a transition from mere participation to active agency (Barroso et al., 2025). This discourse of appropriation signifies the activation of territory as a recognized stake within the social learning space, whereby territory serves as a recognition condition that elevates dialogue from procedural interactions to consequential positioning. Omland et al. (2025) present a more interactional perspective, showing that co-creation occurs through positioning practices in which actors assert, negotiate, or redistribute agency. In this sense, territory becomes the condition for recognition that activates positioning. When participants are granted ownership in the learning environment, they are more likely to engage substantively, articulate their views, and share in the associated risks. Thus, territory connects structural value creation with interactional co-creation, elucidating why dialogue becomes meaningful rather than merely procedural.

Langley et al. (2019) define boundary work as the process through which actors negotiate identities and jurisdictions within organizational contexts. While their analysis sheds valuable light on how boundaries are enacted and transformed, framing territory solely as boundary work may overemphasize identity negotiation at the expense of collective ownership. Boundary work does occur within WIL ecosystems, yet emphasizing territory as ownership underscores the necessary recognition and acknowledgment for sustainable engagement. Boundary work elucidates how positions evolve, while territory explains the rationale behind actors’ investments in these shifts. These relational dynamics are further influenced by time. Gankhuyag et al. (2025) argue that collaborative learning for social sustainability is inherently a long-term endeavour that unfolds over time. Short-term interactions rarely yield deep alignment; instead, ongoing iterations enable participants to recalibrate expectations and foster a shared understanding. Time interconnects with trust and territory, reinforcing relationships and fostering stability. The temporal design of WIL programs, typically structured around 8 - to 12-week collaborative cycles, can determine whether ownership develops or remains superficial.

Additionally, ecosystem complexity emphasizes the necessity for relational anchoring. Jacobides et al. (2018) indicate that ecosystems are defined by distributed agency and interdependence. Governance mechanisms must coordinate functions while allowing for flexibility. In WIL contexts, universities align learning objectives, while private or public organizations contribute domain expertise, and students navigate both. Dialogue, resource accessibility, risk sharing, and transparency do not operate in isolation; instead, they are embedded within the ecosystem’s governance infrastructure. In this conceptual framework, the DART model evolves from a prescriptive tool to a diagnostic lens. Dialogue reflects the quality of boundary-mediated interactions; Access evaluates the availability of resources within social learning contexts; Risk Assessment captures the collective evaluation of vulnerability during collaborative exploration; and Transparency pertains to the visibility of expectations and accountability structures. In WIL settings, these dimensions may empirically shift due to the interdependencies of relational constructs within governance-intensive environments. Therefore, measurement should adapt responsively to contextual variations.

In conclusion, this study defines WIL as a multidimensional learning ecosystem that emerges from boundary-crossing interactions centred on artifacts. Collaboration functions as the generative mechanism for co-creating value, while boundary objects facilitate interaction and stabilize expectations. Interpersonal factors such as trust, time, and territory enhance productive discussions and mutual agency. Territory, understood as ownership in social learning spaces, links structural value creation with interactional co-creation by activating positioning and voice. Collaborative governance frameworks create the macro-architectural conditions that align compliance and capacity building rather than allowing them to conflict. By situating empirical explorations within this theoretically robust framework, we can not only deepen the discussion around measurement in WIL but also contribute valuable insights toward solidifying its theoretical foundations as a distinct field of study.

The present study aims to examine whether the relational architecture of value co-creation, as operationalised through the DART framework (Prahalad & Ramaswamy, 2004; Taghizadeh et al., 2016), retains structural validity when situated within a governance-intensive WIL ecosystem. Rather than treating DART as a universally stable measurement instrument, this study positions its empirical validation as a theoretically informed test of dimensional robustness within a multidimensional learning phenomenon characterised by boundary crossing, stakeholder interdependence, and artefact-mediated collaboration. In this study, DART is used as an analytical starting point rather than a fixed explanatory framework, enabling the empirical assessment of its dimensional stability in context.

As articulated in the Conceptual Framing, WIL is conceptualised here as a governance-structured ecosystem in which universities, students, and external partners co-produce knowledge through boundary-mediated interaction around artefacts. Within such ecosystems, collaboration is not merely coordinative but generative. Value emerges in social learning spaces (Wenger-Trayner & Wenger-Trayner, 2020) through dialogue, positioning, negotiated responsibility, and shared ownership (Omland et al., 2025). These dynamics are sustained by relational preconditions such as trust, time, and territory (Gankhuyag et al., 2025) and are embedded within broader collaborative governance regimes (Emerson & Nabatchi, 2015; Waardenburg et al., 2025).

In this context, the DART framework offers a structured representation of relational conditions for co-creation – Dialogue, Access, Risk Assessment, and Transparency. However, because DART was initially developed in commercial innovation environments, its dimensional distinctiveness cannot be assumed to remain stable when transferred to a WIL ecosystem characterised by pedagogical mediation, distributed accountability, and temporally bounded collaboration cycles. The present study, therefore, investigates whether DART’s theoretical building blocks reorganise empirically in response to the structural and relational conditions specific to WIL.

The empirical setting for this examination is the WE: Working and Envisioning Community of Practice at Bragança Polytechnic University (Instituto Politécnico de Bragança, IPB), where multidisciplinary student teams and external stakeholders engage in 8–12-week challenge-based collaboration. This ecosystem provides a structured yet relationally intensive environment in which compliance frameworks, capacity-building dynamics, and shared responsibility intersect. Within such a governance-intensive learning architecture, co-creation may not manifest as four discrete dimensions; however, it may be reconfigured as relational constellations reflecting how stakeholders experience interaction, ownership, transparency, and shared risk.

Accordingly, the study pursues the following research questions:

RQ1. To what extent does the original four-dimensional structure of the DART model (Dialogue, Access, Risk Assessment, and Transparency) replicate within a governance-intensive WIL ecosystem?

RQ2. If such replication is not observed, what alternative dimensional configurations emerge when the DART scale is applied to external stakeholders participating in boundary-mediated, artefact-centred co-creation processes?

RQ3. What are the psychometric properties – factor structure, reliability, and internal consistency – of the Portuguese adaptation of the DART scale in this WIL context?

RQ4. How can any observed dimensional restructuring be theoretically interpreted in relation to ecosystem governance complexity, social learning spaces, relational trust dynamics, and ownership (territory) within WIL?

By addressing these questions, the study intends to contribute to both measurement theory and WIL scholarship. Empirically, it examines whether established co-creation constructs preserve structural coherence outside their original commercial domain. Theoretically, it tests whether relational mechanisms of dialogue, access, risk, and transparency manifest differently when embedded within governance-intensive learning ecosystems characterised by boundary crossing and shared responsibility. In doing so, the study contributes to efforts to consolidate WIL as a theoretically grounded field rather than a practice-led aggregation of partnership arrangements.

The present study adopts a quantitative, instrument validation research design situated within a post-positivist tradition of construct measurement. The primary objective is not merely statistical replication of an existing scale, but contextual validation of a relational measurement framework within a governance-intensive WIL ecosystem. As articulated in the Aim and Research Questions, the study investigates whether the relational architecture operationalised through the DART framework – Dialogue, Access, Risk Assessment, and Transparency – retains dimensional distinctiveness when transferred from commercial innovation contexts to a boundary-mediated university-industry learning ecosystem.

In measurement theory, the contextual transfer of instruments across domains raises questions about construct stability and dimensional invariance. Constructs validated in one institutional or sectoral configuration cannot be assumed to function identically in another, particularly where governance logics, accountability structures, and relational expectations differ. This is especially pertinent in WIL ecosystems, where co-creation unfolds within pedagogically mediated and temporally bounded collaboration cycles embedded in institutional governance frameworks. The present design, therefore, positions instrument validation not as a purely technical procedure but as a theoretically informed inquiry into the contextual robustness of relational constructs.

Methodologically, the study employs a cross-sectional survey design combined with Exploratory Factor Analysis (EFA). This design is appropriate given the research questions, which seek to examine replication (RQ1), identify emergent dimensional configurations (RQ2), assess reliability (RQ3), and interpret structural shifts theoretically (RQ4). The use of EFA reflects an epistemological recognition that dimensional structures may reorganise in response to contextual conditions. Rather than imposing confirmatory constraints based on the original four-factor DART structure, the analytic strategy allows latent relational dimensions to emerge inductively from stakeholder perceptions within the WIL ecosystem under study.

This section presents the empirical results, structured to align directly with the research questions. The findings are organised into four parts: (1) assessment of factorability and data adequacy; (2) replication analysis of the original four-factor DART structure (RQ1); (3) identification of emergent dimensional configurations and psychometric properties (RQ2–RQ3); and (4) theoretically grounded interpretation of dimensional restructuring within a governance-intensive WIL ecosystem (RQ4). Throughout, results are reported with attention to methodological rigour, construct validity, and contextual coherence.

Prior to factor extraction, diagnostic tests were conducted to assess the dataset’s suitability for exploratory factor analysis. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy indicated satisfactory adequacy (KMO = 0.83), exceeding the commonly accepted threshold of 0.60. Bartlett’s Test of Sphericity was statistically significant (χ², p < .001), confirming that inter-item correlations were sufficiently strong to justify factor extraction.

Inspection of the correlation matrix revealed multiple coefficients above 0.30, indicating meaningful shared variance among items. Communality values after initial extraction were generally above 0.50 for retained items, suggesting that the factors accounted for a substantial proportion of item variance. No extreme multicollinearity issues were observed.

Although the sample size (n = 46) is modest, the analysis was conducted within an exploratory validation logic, and factor loadings were strong and well above minimum interpretive thresholds. Given the high communalities and clear loading patterns, the factor solution can be considered statistically interpretable within the scope of this contextual study. These diagnostics support proceeding to dimensional analysis.

The present study examined whether the original four-dimensional DART framework, namely Dialogue, Access, Risk Assessment, and Transparency, replicates within a governance-intensive Work-Integrated Learning (WIL) ecosystem structured as an artefact-centred social learning space. The empirical results indicate that the original four-factor configuration did not replicate in this setting. Instead, exploratory factor analysis identified a two-factor structure integrating Dialogue with Access and Risk Assessment with Transparency. This discussion interprets these findings in relation to ecosystem complexity, collaborative governance architecture, and WIL as a multidimensional learning phenomenon, while maintaining caution regarding scope and generalisability.

The present study sets out to examine whether the relational architecture of value co-creation, as operationalised through the DART framework (Dialogue, Access, Risk Assessment, and Transparency), retains structural validity when transferred from commercial innovation contexts to a governance-intensive Work-Integrated Learning (WIL) ecosystem. Positioned within a multidimensional conceptualisation of WIL as a boundary-mediated and artefact-centred social learning space, the study did not assume dimensional stability but treated replication as an empirical question.

The findings provide a clear answer to the research questions, subject to appropriate methodological qualifications. Regarding RQ1, the original four-dimensional DART structure was not replicated within the WE governance-intensive WIL ecosystem. Exploratory factor analysis (EFA) revealed that the theoretically distinct dimensions did not yield four statistically independent latent constructs. Attempts to force a four-factor solution resulted in unstable loadings and reduced interpretability. This indicates that dimensional distinctiveness observed in commercial innovation settings is not structurally preserved under the governance and pedagogical conditions characterising the present WIL ecosystem.

Regarding RQ2, an alternative, statistically robust two-factor configuration emerged. Dialogue and Access were integrated into a coherent interactional-operational domain, while Risk Assessment and Transparency converged into a governance-assurance domain. These factors were not arbitrary statistical artefacts but demonstrated strong loadings, minimal cross-loadings, and conceptual interpretability consistent with the structural conditions of the ecosystem under study.

Regarding RQ3, the Portuguese adaptation of the DART scale demonstrated strong psychometric properties in this context. Internal consistency exceeded recommended thresholds (α > 0.80 for both factors), communalities were adequate, and the two-factor solution explained a substantial proportion of variance (72.91%). These results support contextual reliability and preliminary construct adequacy. However, as emphasised throughout the manuscript, this constitutes exploratory validation within a single ecosystem. Confirmatory testing across independent samples and institutional contexts remains necessary before claims of dimensional stability can be advanced.

Finally, addressing RQ4, the observed dimensional restructuring was interpreted considering the conceptual framing of WIL as a governance-intensive, artefact-centred ecosystem embedded in social learning spaces. The empirical convergence of Dialogue and Access is theoretically consistent with boundary-mediated collaboration in which communicative exchange and resource accessibility are enacted simultaneously through shared artefacts. Similarly, the integration of Risk and Transparency aligns with collaborative governance theory, where vulnerability and accountability are structurally interlinked within principled engagement regimes. These interpretations remain inferential: the analysis reveals structural covariance patterns rather than directly measuring cognitive or behavioural mechanisms. Nevertheless, the observed structural patterns are theoretically coherent within the frameworks of ecosystem governance and social learning.

Taken together, the study substantiates the central proposition that co-creation mechanisms may reorganise when embedded within governance-intensive WIL ecosystems. Rather than invalidating the DART framework, the findings indicate that its relational building blocks may manifest differently when situated within pedagogically mediated, temporally bounded, and accountability-structured collaboration architectures. Importantly, the study does not claim universal dimensional restructuring. The two-factor configuration identified here is empirically bound to the WE ecosystem. The contribution lies not in proposing a replacement model, but in demonstrating that dimensional transferability cannot be assumed across institutional logics. Relational measurement instruments require contextual validation when migrating from commercial innovation environments to higher education ecosystems characterised by distributed responsibility and artefact-mediated interaction.

From a theoretical standpoint, the results reinforce the framing of WIL as a multidimensional learning phenomenon rather than a placement-based pedagogical technique. The convergence of relational constructs suggests that collaboration in governance-intensive WIL environments is structurally integrated rather than segmented. Dialogue and access co-function within operational engagement; transparency and risk co-function within accountability architectures. This supports ecosystem and collaborative governance perspectives in which value emerges through coordinated interdependence rather than isolated relational components. From a measurement perspective, the study contributes to construct validation debates by illustrating how exploratory factor analysis can reveal context-specific restructuring when instruments are transferred across domains. The findings underscore the importance of examining dimensional robustness before applying co-creation metrics within higher education ecosystems. From a practical standpoint, the study offers cautious diagnostic insight. In governance-intensive WIL settings similar to the WE programme, relational processes may be more accurately captured at the level of integrated domains – interactional-operational quality and governance-assurance alignment – rather than as four analytically discrete dimensions. However, this suggestion remains provisional and should not be interpreted as prescriptive guidance for instrument redesign without further validation.

Several limitations frame the scope of these conclusions. The study is based on a single institutional ecosystem and a modest sample size. The design is cross-sectional and limited to external stakeholder perceptions. The analysis is exploratory rather than confirmatory. Consequently, generalisability across WIL ecosystems, cultural contexts, and disciplinary domains cannot be assumed. As such, several concrete methodological extensions, including multi-site replication, confirmatory factor analysis, measurement invariance testing, longitudinal designs, and explicit operationalisation of territory and ownership, were outlined in the Future Research Directions section. Building on those proposals, the focus shifts to emphasise the broader research trajectory implied by the present findings.

Advancing the body of knowledge in this area requires moving from context-specific exploratory validation toward comparative and cumulative research. Multi-ecosystem studies can assess whether dimensional convergence is characteristic of governance-intensive WIL more broadly or contingent upon specific architectural features. Integrating student and facilitator perspectives can clarify whether relational restructuring is actor-dependent. A longitudinal investigation can determine whether dimensional integration strengthens or differentiates over time. Finally, explicit measurement of ownership (territory) and psychological safety may illuminate whether these relational preconditions mediate the integration of co-creation constructs.

In conclusion, this study contributes to the theoretical and empirical consolidation of WIL by demonstrating that relational measurement tools must be aligned with the structural realities of governance-intensive learning ecosystems. It advances a context-sensitive understanding of co-creation measurement without collapsing theoretical constructs into reductionist simplifications. By empirically examining dimensional robustness and theoretically situating structural restructuring within ecosystem governance and social learning perspectives, the study strengthens the dialogue between WIL scholarship and measurement theory. The future development of WIL as a field depends not only on innovative pedagogical designs but also on robust conceptual grounding and empirically validated relational instruments capable of capturing how value is co-created across institutional boundaries.

The authors declare that they have no known financial or non-financial competing interests that could have influenced the research, authorship, and/or publication of this article.