Exploring The Utility of Causal Loop Diagrams for Analysing the Continuing Engineering Education Ecosystem

Research Paper Recommended citation: Linnéusson, G., Smith, C., Nizamis, K., & Urenda Moris, M. (2025). Exploring The Utility of Causal Loop Diagrams for Analysing the Continuing Engineering Education Ecosystem. In Kangaslampi, R., Langie, G., Järvinen, H.-M., & Nagy, B. (Eds.), SEFI 53rd Annual Conference. European Society for Engineering Education (SEFI), Tampere, Finland. DOI: 10.5281/zenodo.17631581. This Conference Paper is brought to you for open access by the 53rd Annual Conference of the European Society for Engineering Education (SEFI) at Tampere University in Tampere, Finland. This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License. EXPLORING THE UTILITY OF CAUSAL LOOP DIAGRAMS FOR ANALYSING THE CONTINUING ENGINEERING EDUCATION ECOSYSTEM G. Linnéussona,1, C.J.M. Smithb, K. Nizamisc, M. Urenda Morisd a University of Skövde, Skövde ,Sweden https://orcid.org/0000-0001-8188-7288 Glasgow Caledonian University, Glasgow, Scotland https://orcid.org/0000-0001-5708-6341 c University of Twente, Enschede, The Netherlands https://orcid.org/0000-0002-6965-0242 d Uppsala University, Uppsala, Sweden https://orcid.org/0000-0001-5100-4077 Conference Key Areas: Continuing education and life-long learning in engineering; Dialogue between engineering and society – effects on education (Please select two Conference Key Areas) Keywords: Continuing Engineering Education; causal loop diagrams; systems thinking; education ecosystem ABSTRACT Continuing Engineering Education (CEE) plays a vital role in equipping professionals with the skills needed to navigate technological advancements and sustainability transitions. However, the CEE ecosystem is complex, with multiple stakeholders, interdependent factors, and dynamic trade-offs that challenge effective decisionmaking. This study explores whether systems thinking, specifically causal loop diagrams (CLDs), can provide a structured approach to analysing these dynamics and informing policy and institutional strategies. Using an inductive, qualitative approach, we developed a CLD to map the Swedish CEE ecosystem in the context of the Green Transition. The model highlights key reinforcing and balancing feedback loops that shape the system, including the interplay between competence development, industrial needs, labour market 1 Corresponding Author G. Linnéusson gary.linneus[email protected]e dynamics, and educational adaptation. It reveals how upskilling can drive innovation and economic growth while simultaneously introducing tensions such as workforce turnover, recruitment challenges, and institutional inertia. The findings underscore that effective CEE policy requires system-wide coordination rather than isolated interventions. This study demonstrates the utility of CLDs as a tool for visualising trade-offs, identifying leverage points, and fostering multi-stakeholder dialogue. While the model is exploratory, it serves as a foundation for future participatory validation and refinement. By applying systems thinking, this research contributes to a more integrated understanding of CEE and offers a methodological basis for strategic decision-making in education and workforce development. 1 INTRODUCTION Continuing Engineering Education (CEE) is becoming increasingly vital as it aims to reskill and upskill professional engineers to navigate a rapidly changing world (Cedefop, 2023; IET, 2025). For universities, it serves as a pathway to societal impact (Knudsen et al., 2021; Soeiro, 2006), while for industries, it enhances agility and resilience, empowering employees to adopt new technologies effectively (WEF, 2024). CEE is also essential for smart societies and the digital transformation, driving economic progress, social inclusion and equal access to digital tools. Ultimately, it enables individuals and societies to thrive, innovate, and actively participate in the digital era, making it a cornerstone of smart, sustainable societies. Sustainable societies can further benefit from the transition to green technologies, which are transforming industries by minimizing environmental harm, improving efficiency, and promoting sustainability. Innovations such as renewable energy, electric vehicles, sustainable agriculture, and green buildings help combat climate change, reduce pollution, and preserve natural resources, all while also fostering economic growth, job creation, and improved public health. This transformation demands new competencies from professionals, and CEE must be prepared to meet the challenge. Since the 1980s, SEFI and the International Association of Continuing Engineering Education (IACEE) have addressed and put on the map several enduring challenges CEE still faces today that call for innovative solutions. Among others, curriculum revisions are necessary to keep up with technological advancements (Azofeifa et al., 2024). The industry faces difficulties in combining work and the need to develop employees because of the expenses associated with higher education (Mitra & Raskin, 2023), and the risk of not being able to retain them (Väätäjä et al., 2024). Existing programs offered and the skills needed in the industry are not always aligned (Caratozzolo et al., 2024). Infusing incentive and engagement is very crucial, especially because there is no perceived immediate use of the new skills gained (Väätäjä et al., 2024). The emergence of many certifications makes it difficult to offer credentials and identify where there is a need to embrace different learning styles and teaching approaches. Tackling these issues calls for all stakeholders to work together and co-create across system boundaries. Many of these challenges persevere due to the complexity and dynamic nature of the CEE landscape. CEE constitutes a system of ecosystems with many interconnected variables that efforts to analyse and understand them separately lead to partial fragmented solutions. CEE challenges should be addressed by navigating delicate trade-offs between the different stakeholders rather than optimizing one aspect or another. Recent efforts have been focused on creating a common language to characterize comprehensively CEE and offering a foundation for further discussions (Caratozzolo et al., 2024). However, in this work, the focus is on understanding the dynamics of such a complex system, as no studies in extant literature have been identified seeking to develop a systems-level perspective of Continuing Engineering Education. Systems thinking can provide us with powerful tools to better analyse and understand dynamic complexity in multifactorial systems whose components are entangled and interconnected. To identify these dynamic patterns of complex social system behaviour and dynamic interactions between system components, CLDs (Fig1.) have been proposed and used since 1981 (Kirkwood, 1998; Richardson & Pugh III, 1997)in various application areas, including healthcare, and public policy (Crabolu et al., 2023; Linnéusson et al., 2022). Thus, by applying CLD to the domain of CEE, our study does not claim to introduce a novel methodological tool per se, but rather to explore its value and relevance in a previously under-explored area. Fig. 1. Example of a reinforcing and a balancing causal loop. The left loop is reinforcing (even or zero number of “-”), meaning a change in one direction amplifies itself (e.g., money in a savings account earns interest, increasing the balance and generating more interest). The right loop is balancing (odd number of “-”), meaning a change in one direction triggers an opposing effect (e.g., hunger leads to eating, which reduces hunger). The figure was created using Vensim. Given the lack of existing studies applying CLDs to educational systems and CEE, this study explores the research question: “How clearly can systems thinking provide a method to aid decision-making for impactful co-creation of value in CEE?” Specifically, how can CLDs help explain key dynamics and visualise trade-offs, enablers and barriers within a CEE system? 2 METHODOLOGY The lack of prior studies has resulted in a pragmatic exploratory, multi-method qualitative, inductive approach being utilised; an approach consistent with gaining practical insights to support and inform future practice and encompasses concepts and practices (Saunders, Lewis and Thornhill, 2023, p147). Moreover, as the research question is in essence a methodological question, namely how clearly can CLDs be used to map the CEE ecosystem, then the focus of the paper’s methodology was to limit the scope to determine whether the CLD method had utility in this context, in particular Green Transition in Sweden, as presented in Fig. 2. Open discussions took place between Authors 1 and 4 about the CEE landscape in Sweden, with a particular focus on Green Technologies and the Green Transition. These discussions centred on Author 4’s understanding of the system, being an owner-of-question at a larger University for developing their CEE ecosystem. Subsequently, Author 1 inductively coded the transcripts and notes to identify key actors and their interactions. Through multiple iterations, this process resulted in a version of a CLD that reflected the initial understanding. Key factors were also derived from an earlier study proposing a taxonomy for CEE practices (Caratozzolo et al., 2024). The CLD was then presented and refined through several online discussions with all authors. Once all authors agreed that the CLD had sufficient clarity, its utility was evaluated. This approach aligns with established methods for developing CLDs (Baugh Littlejohns et al., 2021). The utility of the model was addressed in two ways, firstly an internal verification by author 4 (who drove the initial creation process), and then a validation by author 3 who was not involved as much in the early development. Fig. 2. A graphical representation of the methodology the authors followed for creating (Data Collection), refining, verifying, and validating the proposed CLD, in that order. The verification process was conducted by Authors 1 and 4 to ensure that the model accurately reflected the experiences, perceived reality, and biases of the initial description, as well as to validate the feedback loops. The process followed this approach: each loop was described and analysed by Author 4. Special attention was given to causal loops that modelled critical, contrasting goals and those where the perceived output led to unexpected results. Additionally, names and outputs were verified and further discussed. The outcome of this process was an extended model description, which served as the basis for the explanation presented in Section 3. After the model verification, Author 3 performed a single-person validation. This process included the identification of all causal relationships and a check on the assumptions that dictated each causality. The purpose was to check if those are generalizable or specific, if the model can be simplified without loss of information, and if the loops identified are consistent with reality. Additionally, it identifies hidden loops and assesses the polarity of the causal relationships. Additionally, he checked if the diagram sufficiently represents what happens in practice. 3 FINDINGS 3.1 The Current CLD of the Swedish CEE Ecosystem for Green Transition The CLD illustrates how pressures related to the green transition drive complex interactions between companies, education systems, and macro-level policy environments. Competence development and upskilling support productivity gains and technological progress, but also increase demands on the labour market and educational capacity. Technological transitions and economic growth create new opportunities while simultaneously introducing systemic tensions, such as recruitment challenges, turnover, and institutional inertia. The education ecosystem emerges as a central actor, balancing responsiveness with internal limitations. Fig. 3. CLD of the Swedish CEE ecosystem for Green Transition, using Vensim. need for green transition technologies in use + technological shifts apt edu CEE edu offerings in use quality of edu offerings productivity competence in companies workload - + - +need for competence domains in use + + - + employee turnover + + investments job market possibilities + available apt employees to hire - + governmental fundings - + + green tech developments - active industrial businesses + pressure to develop green transition competence + + B1 Societal need Short-term Edu response B7 Competence reinvestment R2 B3 R3 Transition in competence Transition in workload R4 Transition in Industrial job market + + + + + - resource constraints in edu systems + - apt edu programs + - + + + attractiveness of edu programs financial status of edu ecosystem - + Competence loss + time and competence to consider relevance to society - + Stimulation of Industrial ecosystem Stimulation of Edu ecosystem Competence availability Long-term Edu response - + + R7 R5 External attractiveness Edu R11 Internal attractiveness Edu Transition in Edu need Economic growth through technological shifts R9 R6 B10 R10 + B2 Industrial response R1 B5 R14 R13 B9 B8 Transition in technologies Government policies Legend for colours: Macro Environment Industrial Ecosystem Green Transition Need Educational Ecosystem Refinement of Edu offerings New Edu offerings R12 Unintended consequences Edu workload mobility of competence + + + B4 Transition in the Macro job market Worker mobility Regional mobility magnification B6 R8 + R B Reinforcing loop; arrow indicates directionality Balancing loop; arrow indicates directionality The current Swedish CEE ecosystem is visualised in Figure 3 through a CLD. As described in the legends, the model is structured around five key subsystems according to colour themes: Green Transition Need, Industrial Ecosystem, Macro Environment, Government Policies, and Educational Ecosystem. However, in the description below, we narrate the CLD using four key thematic areas to explain the feedback dynamics shaping the Swedish CEE system for green transition. 3.2 Description of the CLD We recommend that readers closely follow the CLD while reading the description, where each loop—balancing (B) or reinforcing (R)—is clearly identified to support understanding. 1. Competence Development as a Response to Societal Need (B1, B2) – As the demand for a green transition intensifies, societal expectations grow for companies to develop relevant competencies (B1). In response, companies invest in upskilling, aiming to increase productivity and support technological advancements that enable green solutions to replace outdated ones, thereby easing the urgency of transformation (B2). 2. Transitions and Tensions in the Industrial Ecosystem (R1-R7, B3-B5) – Improved productivity gains can reduce workloads, enabling companies to reinvest time into employee competence development, fueling a virtuous cycle of knowledge and performance (R1). However, accelerating the use of emerging technologies during transitions can multiply workloads and increase employee turnover, introducing friction in the system (B3, R2, R4). These tensions may intensify as job market opportunities expand, driven by investments in the macro level (R7, B4), reshaping labor dynamics (B5). Over time, these transitions render both existing competencies (R3) and technologies (R5) obsolete, dissolving earlier pressures. To drive the desired transition, the industrial ecosystem is synergistically co-dependent on the adaptive capacities of the educational ecosystem (R6). 3. Macro Environment Drivers of Technological Transitions (B6, R8-R11) – As a strategic response to stimulate the desired technological transition, governmentfunded macro-level investments in regional development can help rebalance regional mobility (B6), improving competence mobility and easing the recruitment of critical competences (R8). In the short term, macro-level funding can accelerate technological progress (R9) and generate system-wide economic growth (R10). However, as shown by loops R4, B5, and R11, these targeted interventions may also unintentionally intensify job market pressures, amplifying employee turnover within the industrial ecosystem, hindering recruitment, and potentially reducing interest in education among certain groups. 4. Education Ecosystems Adaptation to Societal Pressures (B7-B10, R12-R14) – In response to pressures for green transition competencies, education ecosystems must evolve their offerings. Universities may introduce agile CEE programs to meet immediate industrial needs (B7) and develop long-term offerings, though often slowed by institutional inertia (B8). The ability to generate relevant offerings depends on financial stability and existing quality across the education ecosystem (B9). However, rapid expansion risks introducing internal resource constraints and overloaded workloads, leading to unintended quality reductions (R12). Internal attractiveness of the education system (R13) can reinforce either virtuous or vicious dynamics, depending on system conditions and the capacity to prioritize societal relevance (R14). Therefore, when resource constraints increase, foresight capacity may erode, leading to misalignment between education offerings and long-term societal needs (B1 and R14). These internal dynamics are reinforced by pressures from macro-level interventions (R7, R9, R11), highlighting the importance of investing in institutional foresight to refine offerings in line with societal, not just market, needs (R14). This highlights how governmental interactions with the education ecosystem, through funding and incentives, may influence systemic balance, depending on how they interact with existing institutional dynamics (B10). 4 DISCUSSION AND CONCLUSIONS This study set out to explore whether systems thinking—and in particular, causal loop diagrams (CLDs)—can support decision-making and insights in the context of Continuing Engineering Education (CEE), especially amid the pressures of a green transition. As indicated herein, where CLDs have proven valuable in other contexts, the resulting model in this study similarly offers a structured view of the complex interplay between system elements, here competence development, societal needs, technological transitions, labour market dynamics, the education system, and government policy interventions. 4.1 Insights from the Model The CLD reveals how reinforcing loops—such as those between upskilling, productivity, and technological innovation—interact with balancing and constraining forces, including workload pressures, institutional inertia, and labour market volatility. Industrial transformation is not solely a matter of increasing educational output; rather, the model highlights the critical role of system alignment and coordinated adaptation. The educational ecosystem emerges as both an enabler and a potential bottleneck, depending on its internal foresight capacities, adaptiveness, and quality. The analysis also shows that macro-level interventions, while often intended to stimulate growth and innovation, can inadvertently amplify internal tensions within industrial and educational subsystems. This suggests that policy design for CEE should consider not only where to intervene but also how systemically those interventions interact with existing dynamics. 4.2 The Value of CLDs in CEE Research and Practice This study demonstrates the value of applying CLDs in CEE by highlighting their capacity to enhance understanding of complex dynamics and support more open and collaborative decision-making. From a research perspective, the CLD provided a way to surface and visualise the dynamic interdependencies shaping CEE ecosystems, making explicit how competence development, technological transitions, labour market dynamics, education system adaptation, and macro-level policy interventions interact over time. This explicit mapping supports researchers in identifying leverage points, feedback structures, and tensions that may not be apparent when studying isolated system components. Considering the value to practice, by its visualisations, the CLD can be argued to help foster deeper understanding among stakeholders by enabling them first to “see” the system — identifying key variables and their relationships — and then move towards a more profound “understanding” of how different actions and interventions might shape outcomes. This process of visualisation and reflection can support stakeholders in transforming their mental models, a critical step towards developing more holistic, systemic solutions in CEE. By facilitating the surfacing of mental models, making them more explicit and shared, the CLD becomes a tool for dialogue, co-creation, and strategic thinking, creating the ground for more open decision-making that aligns with the complexity of the CEE landscape. Thus, using CLDs in CEE research and practice contributes to both improved system awareness and a shared foundation for collaborative, strategic co-actions. 4.3 Methodological Reflections and Limitations One challenge encountered in developing the CLD was balancing ‘accuracy’ (correct cause-effect structure, grounded assumptions) with ‘usefulness’ (stimulating discussion, promoting understanding, guiding strategic thinking). While this model aimed for structural validity, it prioritizes insight over precision. The model was developed from a single, experienced source with multi-stakeholder insight and validated internally. It is not yet empirically generalized but provides a strong conceptual basis for stakeholder engagement and future research. 4.4 Future Work and Contribution The next step is to use the CLD in participatory workshops with multiple stakeholders to test and explore the model, considering its usefulness as a boundary object across positions in the CEE ecosystem to generate actionable insights. This process will not only enhance the model’s validity but may also establish it as a co-creation tool for policy and curriculum design. Beyond this next step, future work will explore the CLD’s scalability properties within other CEE contexts in other European countries and explore possibilities to further combine this approach with system dynamics simulation or other system methods to quantitatively support exploration of trade-offs between short-term and long-term impacts of various interventions. REFERENCES Azofeifa, J. D., Rueda-Castro, V., Camacho-Zuñiga, C., Chans, G. M., MembrilloHernández, J., & Caratozzolo, P. (2024). Future skills for industry 4.0 integration and innovative learning for continuing engineering education. 9. https://doi.org/https://doi.org/10.3389/feduc.2024.1412018 Baugh Littlejohns, L., Hill, C., & Neudorf, C. (2021). Diverse approaches to creating and using causal loop diagrams in public health research: recommendations from a scoping review. Public Health Reviews, 42, 1604352. Caratozzolo, P., Smith, C. J., Gomez, S., Moris, M. U., Nørgaard, B., Heiß, H.-U., & Azofeifa, J. D. (2024). A novel taxonomy for facilitating in-depth comparison of continuing engineering education practices. 9, 1444595. https://doi.org/https://doi.org/10.3389/feduc.2024.1444595 Cedefop. (2023, November 28). Researchers and engineers: skills opportunities and challenges (2023 update) | CEDEFOP. https://www.cedefop.europa.eu/en/datainsights/researchers-and-engineers-skills-opportunities-and-challenges-2023-update Crabolu, G., Font, X., & Eker, S. (2023). Evaluating policy complexity with Causal Loop Diagrams. Annals of Tourism Research, 100, 103572. https://doi.org/10.1016/j.annals.2023.103572