Benchmarking Project-Based Learning Practice Internationally

Research Paper Recommended citation: Goldfinch, T., Belkina, M., Brown, N., Miao, G., & Grundy, S. (2025). Benchmarking Project-Based Learning Practice Internationally. 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.17631865. 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. BENCHMARKING PROJECT-BASED LEARNING PRACTICE INTERNATIONALLY T Goldfinch a,1, M Belkina b, N Brown c, G Miao d, S Grundy e a University of Sydney, Sydney, Australia, 0000-0001-5506-136X b University of New South Wales, Sydney, Australia, 0009-0006-2660-2845 c RMIT, Melbourne, Australia, 0000-0002-2562-2258 d University of Technology, Sydney, Australia, 0000-0001-6912-0391 e University of New South Wales, Sydney, Australia, 0009-0009-9018-7385 Conference Key Areas: Curriculum development and emerging curriculum models in engineering; Building the capacity and strengthening the educational competences of engineering educators Keywords: Project-based Learning, Benchmarking ABSTRACT Project-based learning (PjBL) is a commonly used approach to developing students’ transferrable skills, familiarity with professional practice, and design skills among other higher order capabilities. While much has been published on the methods and outcomes of PjBL in recent decades, studies tend to focus on best practice casestudies or particular educational foci of the studied approaches. Understanding the range of current practices, particularly in terms of practicalities of implementation such as resourcing, assessment structure and weighting requires extensive literature review or, more commonly, informal benchmarking of practices between colleagues. Recognising this gap, the authors of this paper developed an approach to benchmarking PjBL practices across institutions. The approach was implemented as an Australian national benchmarking project in 2021-22 and has since been awarded for its research design by the Australasian Association for Engineering Education. The team has adapted the method to support a global benchmarking of PjBL practice. The revised method is presented here alongside sample data from the Australian study to illustrate its utility. The authors invite colleagues to participate in broader data gathering to understand how their practices compare internationally. 1 Corresponding Author T Goldfinch [email protected] 1 INTRODUCTION Project-Based learning (PjBL) is an educational approach commonly used to engage students in authentic learning experiences that have practical industry relevance (Blumenfeld et al. 1991; Herrington & Herrington 2006) and is defined by its central premise of a problem in the form of a project. While PjBL is not new it is consistency cited as an important tool to develop graduates who are industry ready and adequately prepared to contribute to society, including its ever-evolving requirements of engineers (Kolmos and de Graaff 2014; Lawrence 2020). The reported benefits of PjBL are well established and include the development of generic or professional skills (Boelt et al 2022; Picard et al 2021), identity (Ju & Zhu 2023), learner capability (Nguyen et al 2020; Zhang & Ma 2023), interdisciplinarity (MacLeod & van der Veen2019; Van den Beemt et al 2020) as well as providing a better structure/framework for developing understanding of the complexities of engineering practice and how to apply knowledge in the real world (Mills & Treagust 2003; Bauters et al. 2020; Evenddy et al. 2023). Despite the benefits, there are significant challenges to wider adoption of practicebased approaches such as PjBL which are reported globally. These include: perceived resource-intensity and the cost associated with scaling up projects for large cohorts (Miao et al 2024b); access to industry project partners; appropriately qualified or experienced teaching staff (Mitchell & Rogers 2019); challenges of engaging diverse cohorts; and, workload models which favour traditional teaching approaches (Crosthwaite 2021). These challenges can result in implementations of PjBL which stray from the intended authenticity of the mode, impacting students’ experiences of realistic, authentic practice within the curriculum. They also feed into workload concerns for educators who use PjBL which can lead to staff attrition in these types of units (Brown 2020; Kavanagh et al. 2012). Given calls for expansion of PjBL in several engineering education change agendas (Graham 2018; Crosthwaite 2019; EWB UK, 2024), our multi-institutional team set about gathering data on practical implementation of PjBL in the Australian context, guided by the following research questions: • What are the current practices and challenges in current PjBL practice? • What are the barriers and enablers of current PjBL practice to meet the Australian Council of Engineering Deans Engineering 2035 agenda? To address these questions, the team gathered desktop data from Australian & New Zealand universities to compare and share experiences of PjBL, as well as build an understanding of the support provided to the educators of PjBL units. Findings of the Australasian study have been published previously (Miao et al 2024a; Miao et all 2024b). The method has now been adapted for an international context and is reported here to elicit discussion on the approach, interest in participation, and demonstrate the value of the data collected. 2 METHOD The benchmarking study employs a mixed-methods approach to provide both a broad overview of the project-based learning (PjBL) landscape and a deeper understanding of the experiences of PjBL stakeholders. The benchmarking framework was developed, and data were collected and analysed across Australian and New Zealand universities during 2021–2023. In 2024–2025, the framework was discussed with international collaborators and subsequently adopted for global benchmarking. The approach accommodates various applications of PjBL from a disciplinary perspective, as well as those involved in teaching delivery. The structure of data sources is illustrated in Figure 1. Figure 1. Summary of research components and data sources. 2.1 Desktop data collection The first component involves a desktop study of data on PjBL Units of Study. Units of Study are identified as PjBL by faculty leadership of participating Institutions. The researchers then gathered data directly from Unit of Study coordinators through email correspondence and a simple .pdf form. In the previous study (Miao et al 2024a; Miao et all 2024b), this approach enabled us to identify and clarify categorise current practices based on course design philosophy and format, project focus, delivery and assessment modes, staffing ratios, staff experience/background, industry and community engagement and outcomes of student satisfaction survey. Statistical methods were used to identify not just the statistical significance of relationships in the data, but also effect sizes (i.e. the practical implications of the relationships). Specific components of the data collection are described in Table 1. The use of a .pdf form for this purpose gave respondents flexibility to enter data in the form or scale in which it is available locally along with explanatory notes as required. This data was then interpreted and rescaled manually to enable cross comparison. As is common with benchmarking studies, indicators were limited to those which apply to most implementations of PjBL. This is particularly the case with evaluation data, where other approaches may be used, but student evaluations of teaching (SET) were the only indicators common across most institutions. Table 1. Desktop benchmarking data indicators University context Unit of Study design and resourcing (desktop data) Disciplinebased Units Units taught accross engineering disciplines Units taught with nonengineering students Qualitative semi-strcutured interviews PjBL Unit Coordinators (educators) Industrybased PjBL Collaborators Element Data source Indicators Educational focus Form Online Unit of Study information • Year level • Single/multi-disciplinary • Compulsory/elective study requirement • Technical/transferable skills emphasis Assessment Online Unit of Study information • Technical/transferable skills assessment emphasis • Assessment modes • Assessment weighting Resourcing Form • Academic staff experience • Academic staff seniority • Hours committed to run the unit • Size/shape of teaching team Evaluation Metrics Form • Student evaluation of teaching results 2.2 Interviews The second component involved interviews with seven Unit of Study coordinators and six industry-based PjBL collaborators. We drew upon the data analysis and findings from the desktop study to identify key areas to explore in more detail through these interviews. These interviews aimed to capture deeper insights into the practical implementation of PjBL, challenges faced by educators and collaborators, and the perceived effectiveness of these initiatives in real-world settings. Inductive thematic analysis of interview transcripts was undertaken to identify key areas that impact the design, delivery and implementation of PjBL. Both the desktop study and interviews spanned a range of institution types from large, urban, comprehensive, research-intensive universities (‘Group of 8’ or ‘Go8’ in the Australian context) to smaller regional universities. The study in its entirety (as well as subsequent adaptations for the international context) was reviewed and approved by the University of New South Wales Human Research Ethics Committee. 2.3 Adaptations for Global Benchmarking Expansion of the Australasian study to an international context has required careful adaptations to the language and data capture format for the desktop component. Where the Australian study captured data through a simple .pdf form, the international study has been adapted to a Qualtrics-based online form. Generalisation of language and more detailed explanatory notes on key terms/fields have also been added based on experience from the Australasian study with a view to creating a data collection tool that is less dependent on reinterpretations or follow up clarifications by the researchers. The form remains in English language only. At the time of writing, the revised form has approved by the ethics committee and is currently being distributed through the authors networks. It is expected that a final, live version will be published ahead of the SEFI2025 conference and early results on global benchmarking will be available. Options for follow up interviews and supplementing form data with review of Unit of Study information available online remains unchanged from the original study. 3 RESULTS The findings of this work to date have presented a useful snapshot of contemporary norms in PjBL within the Australian & New Zealand context, as well as insights on next steps of how to sustain and scale PjBL use in engineering curricula. Results presented in this section represent data from 66 PjBL Units of Study across 15 Australian and New Zealand universities. These sample results are provided for the purpose of illustrating the utility of the method and the results which can be expected from a global study, and have been published in expanded form elsewhere (Miao et al 2024a; Miao et all 2024b). Updated results including global data are due to be presented alongside the paper in September. Although there are some degrees internationally that are entirely PjBL, it is far more typical to find intermediate models where PjBL is used in key points throughout a program. There is variety in the focus of projects that are used, the rationales for using PjBL and approaches to PjBL. Moreover, despite interest in development of interdisciplinary capability through PjBL in the international context (Habbal et al. 2024; MacLeod & van der Veen2019; Van den Beemt et al 2020), our Australasain study sample identified only two out of 66 units (at the University of Sydney and the University of Melbourne) where non-engineering students were involved. Both were elective units (optional to study) run in conjunction with a business unit. PjBL is used by academics across all levels; however, as shown in Figure 2, junior academics (Levels A–C) have greater student numbers than senior academics (Levels D–E). The tendency for larger, junior units to be coordinated by junior staff appears to be a widespread and long-standing occurrence across Australian & New Zealand universities. This is concerning as is it has been reported elsewhere that junior staff in engineering faculties often feel powerless to change teaching culture at their institutions (Kavanagh et al. 2012) and often don’t have the agency to seek appropriate resourcing. Concentration of these larger units among more junior staff could also be indicative of their status among faculty. Coordinators have highly variable experience in industry (Figure 3). Whilst 33% of coordinators have no industry experience, 46% of coordinators have significant industry experience (>5 years). Most units in this study (n = 50) contained projects based on real-world problems that are determined by the coordinator. Of these units, 29 were run by a coordinator with less than 5 years of industry experience, highlighting a need to have greater direct involvement of industry partners to ensure contemporary authenticity of projects and engineering practices used within them. Interviews with industry indicated strong interest to be involved, but that institutional barriers and personal constraints can hinder regular engagement. Figure 2: Coordinator level with class size for all units (n = 75) Figure 3: Coordinators’ industry experience (n = 52) Our approach also provided useful data on resourcing such as total unit workload hours (coordinator + sessional tutoring staff). On average, resourcing of 3.5 hrs per student is allocated for PjBL units. This is not to say that this is appropriate resourcing, but it provides a benchmark to which individuals can evaluate their own resourcing arrangements. The mean lead unit coordinator (academic staff member leading the design, teaching and delivery of the unit) workload was equivalent to 0.4 Full-Time Equivalent (FTE); however, some coordinators reported workloads of >1FTE due to class size, highlighting staffing sustainability risks reported elsewhere as barriers to PjBL. In interviews, workload was one of the most common concerns raised by coordinators. Shared unit coordination models were not common (n = 8), but were highlighted as a key approach to managing contingency, reducing workload concerns and increasing psychological safety. On educational focus, the method provided insights on learning emphasis of Unit learning outcomes - technical learning outcomes and professional learning outcomes, coded against the Engineers Australia Stage 1 Elements of Competency (EoC). On average, 60% of a PjBL unit’s learning outcomes are technical; however, this varies along the length of a degree. First-year units typically have the lowest proportion of technical outcomes (39% on average). The proportion testor exambased assessments tended to increase with the proportion of technical learning outcomes, see Figure 4. Figure 4: Technical learning outcomes vs test-based assessments (n = 65) Figure 5: Student satisfaction for multiand single-disciplinary units (n = 21 for single2017-19; n = 11 for multi2017-19; n = 35 for single2020; n = 20 for multi2020) We also found that units with students from multiple disciplines of engineering had lower student satisfaction than single disciplinary courses on average (Figure 5). These units also tended to have a greater focus on professional or transferrable competencies, corresponding with other research indicating student resistance to the complexity of project work which emphasises these skills (Goldfinch et al. 2019; Willey & Matchet, 2018). Having benchmarking data which illustrates this can help unit coordinators and educational leaders to better interpret student evaluation scores in context. 4 Discussion & Conclusion While PjBL is embraced in various forms across the sector, its sustainable and scalable implementation remains highly dependent on institutional support. The data generated from this benchmarking approach supports unit coordinators and faculty leaders in better understanding key elements of their practice, resourcing, and education focus. Institutionally, PjBL units are often compared to units that use different educational approaches, which may not result in favourable outcomes – particularly with respect to staffing resources required and student satisfaction survey scores (Goldfinch et al 2019). The findings from this research allow for a more accurate like-for-like comparison. The work has confirmed and consolidated findings from a range of publications on individual experiences of PjBL (e.g. Tse and Bona 2019; Brown 2020) and provided evidence of shared challenges amongst PjBL offerings that cannot be addressed by unit coordinators individually. It has also provided quantified insights into sustainability and scalability of PjBL, particularly with respect to scaling for large class sizes, providing tangible data and recommendations to support decision-making around academic support. Moreover, the results provide critical support for career progression of individual PjBL unit coordinators, enabling them to construct teaching impact cases for promotion which are benchmarked against comparable Units of Study. Coordinators can use these results to champion appropriate resourcing and recognition of PjBL and advocate for changes that improve practice. The sustainability of PjBL must be also actively addressed by faculty leadership. Key recommendations for removing barriers to expansion of PjBL which have arisen from this work to date have been delivered to engineering education leadership groups in Australia and New Zealand. These include: • Providing fit-for-purpose resourcing models for large and complex PjBL units; • Allowing sufficient time for unit design and industry relationship-building; • Empowering academics with autonomy to create project units suited to their students; • Adopting co-coordination models to distribute workload; • Building shared frameworks for interpreting student feedback that reflect the unique demands of PjBL. These recommendations are significant in multiple ways. For coordinators, they offer mechanisms to sustain PjBL delivery and improve student outcomes. For T&L leadership, they suggest actionable strategies to embed PjBL into broader program structures while supporting staff wellbeing. Sustainability is highlighted as a precondition for scalability; without resolving systemic issues though collation and presentation of robust benchmarking data, the expansion of PjBL across programs and institutions will be limited. In summary, the method presented here and our results to date addresses a significant gap in the engineering education discourse, by benchmarking PjBL practices, with respect to student numbers, assessment types, resource allocation (academic workload and cost per unit), industry engagement and student experience. The results do not (and were not intended to) define ‘best practice’ definitively – rather they provide individual unit coordinators and education leaders with the data needed to understand where their practices are positioned relative to what is common elsewhere. The significance of this research lies not only in the value of the data and findings themselves, but also in the usefulness of the data in empowering PjBL academics to seek enhanced support. This is clear in the continued interest from academics in contributing to the expansion of this snapshot of PjBL and broadening this work to compare Australian and New Zealand PjBL with international approaches. REFERENCES Blumenfeld, P. C., Soloway, E., Marx, R. W., Krajcik, J. S., Guzdial, M., & Palincsar, A. (1991). Motivating project-based learning: Sustaining the doing, supporting the learning. Educational psychologist, 26(3-4), 369-398. Bauters, M., Holvikivi, J., Vesikivi, P. (2020). An Overview of the situation of projectbased learning in engineering education. SEFI Annual Conference 2020. Enschede, Netherlands. Boelt, A. M., Kolmos, A., & Holgaard, J. E. (2022). Literature review of students’ perceptions of generic competence development in problem-based learning in engineering education. European Journal of Engineering Education, 47(6), 1399– 1420. https://doi.org/10.1080/03043797.2022.2074819 Brown, N. (2020). Practical solutions to manage staff and student workloads in project-based learning courses. Global Journal of Engineering Education, 22(1), 2025. Crosthwaite, C. (2019). Engineering Futures 2035: A scoping study. Australian Council of Engineering Deans. Crosthwaite, C. (2021). Engineering Futures 2035 Engineering Education Programs, Priorities & Pedagogies. Australian Council of Engineering Deans. Engineers Without Borders UK & Royal Academy of Engineering (2024). Reimagined Degree Map: A guide for educational providers to keep degrees relevant for engineers graduating before 2030. Retrieved 8th September 2024 from https://www.ewb-uk.org/reimagined-degree-map/. Evenddy, S., Gailea N., Syafriza (2023). Exploring the Benefits and Challenges of Project-Based Learning in Higher Education. PPSDP International Journal of Education Volume 2 (2) 05-06 July 2023, 458-469. Goldfinch, T., J. Vulic, E. Leigh and K. Willey (2019). Student Perceptions of Complexity in Engineering Education. SEFI Annual Conference 2019. Budapest. Graham, Ruth (2018). The global state of the art in engineering education. Massachusetts Institute of Technology, Cambridge, MA