Enhancing Assessment and Feedback in Engineering Education Through Digital Innovation

Research Paper Recommended citation: Jones, E., Riazifar, N., & Watts, K. (2025). Enhancing Assessment and Feedback in Engineering Education Through Digital Innovation. 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.17631220. 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. ENHANCING ASSESSMENT AND FEEDBACK IN ENGINEERING EDUCATION THROUGH DIGITAL INNOVATION E.S. Jones 1, N. Riazifar b, Kim Watts c a University of Warwick, Coventry, UK 0000-0002-1189-1692 b University of Warwick, Coventry, UK 0009-0006-2563-3922 c University of Warwick, Coventry, UK 0009-0002-7593-9955 Conference Key Areas: Teaching mathematics and physics in engineering education Keywords: active learning, online assessment tool, student engagement, automated feedback ABSTRACT Active learning is an important aspect of engineering education enabling the application of STEM subjects to real-world contexts. As digitalization advances and more assessments move online, tutors often struggle to observe the support needs and learning progress of students. Additionally, opportunities for student interaction in online assessment are limited by the capabilities of educational technology. This study provides an adaptable model for integrating technology-enhanced assessment in engineering education. It aims to explore and implement practical changes in assessment design, utilising innovative technologies to enable a supported student experience. Through analysis of student feedback and assessment outcomes, this research investigates the effectiveness of a Moodle integrated digital tool, STACK. The study revealed a positive influence on self-directed learning through the provision of a personalised feedback mechanism. The results highlight a positive correlation between student outcomes and their engagement with the tool and supporting resources. 1 Corresponding Author E.S. Jones [email protected] . 1 INTRODUCTION The requirement for engineering students to apply STEM subjects such as maths and physics to real-world problems means that the discipline naturally lends itself to an active learning approach (Lima, Andersson, & Saalman, 2017). Active learning techniques including problem-, project-, and challenge-based learning, are increasing in popularity (Sukackė et al., 2022), and research findings indicate that active learning can increase student performance (Freeman et al., 2014; Johnson et al., 2025; Ting et al., 2023). The traditional, one-size-fits-all lecturing approach struggles to accommodate an increasingly diverse student population in the engineering disciplines, who have heterogenous levels of knowledge in subjects such as maths (Knaut et al., 2022; Kraska, 2022; Paiva et al., 2015). Instead, educators are adopting strategies that foster self-efficacy in their students such as constructivist learning and self-direction. Mejeh, Sarbach, & Hascher, 2024) argue that self-regulation is a key skill in twentyfirst century working and learning – and that students who can direct their own learning perform better academically. A key question emerges: how can self-efficacy be developed in weaker students? (Saadati et al., 2015) argue that the cognitive apprenticeship theory (CAT) (Collins et al. 1991) supports novices to becomes experts via examples provided by a more knowledgeable practitioner. As Moradi et al., (2018) explain, in CAT, the novice is not only presented with the correct problem-solving method but is also helped to understand the process and decisions that led to the chosen strategy via a modelling process demonstrated by the expert. The novice is led to mastery of the subject via a stepped down support approach of modelling (as described above), coaching (practising observed methods with input from the expert) and fading (where the expert reduces instruction as the novice approaches mastery). However, monitoring students growing mastery in STEM education is a challenge in the coaching phase when students undertake self-directed learning remotely and online. As a result, a study in online assessment and feedback methods in STEM education was proposed to ascertain a suitable coaching mechanism for self-directed, online learning. A range of novel approaches to assessment in maths and engineering have emerged since the forced move to online teaching and learning due to the Covid-19 pandemic. For example, Shi & Shahbazi (2024) and Valdez & Maderal, (2021) reviewed the use of polls and multiple-choice quizzes to check student understanding during synchronous online instruction. One tutor in Shi and Shahbazi’s study lamented that there was no way to know how well students ‘connected’ with the teaching via these methods. And Valdez and Maderal point out that they could not rule out guessing as a means of answering the questions. As Bach2020) notes, multiple-choice quizzes can only test superficial learning. Such tools do not replicate the coaching stage of CAT used in the classroom. Two studies of online maths assessment which were informed by CAT were Moradi et al. (2018) and (Shih-Che & Hung-Hsu, 2022). Moradi et al. provided modelling via video instruction but used multiple-choice questions to test understanding and only provided performance feedback at the end of the instructional module – meaning the coaching and fading stages were not present. By providing a mechanism by which handwritten notes could be shared via the cloud, Shih-Che and Hung-Hsu were able to set much more sophisticated assessment tasks and therefore provide more targeted feedback, but again, this feedback was focussed on error correction after the fact rather than replicating the coaching stage. Crisp, Guàrdia, & Hillier2016) assert that there is a need to explore novel approaches in online assessment in maths. Online assessment which provides coaching in selfdirected online learning is required. Coaching is the stage of the CAT model which moves students towards mastery. This paper explores the research question: how can a digital assessment tool incorporate the coaching stage of CAT to support selfdirected learning in engineering maths education. We investigate the potential of STACK as a solution to bridge the coaching gap in the online environment. Our study adopts a distinct approach by aligning the CAT's model phases with STACK's feedback features. STACK is an online assessment system designed for mathematics and STEM subjects (Ahmed & Seid, 2023; Juma et al., 2022; Knaut et al., 2022; Zerva et al., 2022). Unlike traditional multiple-choice formats, STACK allows students to submit answers in the form of mathematical expressions, such as polynomials. This functionality means that students can answer much more sophisticated questions ensuring a more authentic assessment of their understanding (Kraska, 2022; Paiva et al., 2015). During each question attempt, STACK provides a validation stage in which students are coached to understand the type of answer required. Students have the opportunity to check that their answers can be interpreted by the system. Invalid responses, such as those with mismatched brackets, are highlighted before the ‘correctness’ of the answer is checked. This ensures that students are not unfairly penalised due to formatting errors or minor input mistakes (Zerva et al., 2022). In addition, multipart questions with follow-through marking are supported, allowing students to receive appropriate credit even if an error propagates through different parts of a solution. STACK can provide specific feedback tailored to each student's response and the range of available question types supports weaker students in particular (Bach, 2020). STACK quizzes can be configured to provide instant or deferred feedback, or a combination of both, a function which supports the faded support recommended in CAT (Zerva et al., 2022). Instant feedback is highly valued by students (Ahmed and Seid, 2023), (Juma et al., 2022) and the value of retrying and practising as much or as little as needed is supported by research on mastery learning (Cameron, 2022) (Zerva et al., 2022). Knaut et al. (2022), Cameron (2022) and Kraska (2022) emphasise the benefits of self-directed study provided by STACK which complement its personalised feedback and interactive test and learn functionality. A review of the literature has shown that STACK can fill the coaching gap left by standard online Maths provision by creating an online space in which students can interact with complex maths problems, receive instant feedback and refine their approach through personalised coaching. 2 METHODOLOGY This study examines a pilot implementation of STACK within an undergraduate degree apprenticeship Maths module, Applied Maths and Statistics, delivered in a blended learning format, conducted in compliance with university ethical regulations. STACK was integrated into the module’s Moodle page as an online assessment and feedback tool, embedded within the Moodle quiz function. The assessment comprised three distinct question types developed by the authors: numerical questions in which students inputted calculated data into tables directly; STACK questions, such as maths formulas which required the use of basic syntax to generate mathematical notation and equations; and essay-style questions that combined text, symbolic and numerical input to allow students to provide interpretation and reasoning for their answers. STACK syntax, a lightweight coding format, facilitates the structured representation of mathematical notation within the assessment environment. The categorisation of these question types aimed to evaluate how different formats influence students' problem-solving approaches, conceptual understanding, and engagement with mathematical notation in an online setting (Bach, 2020). A comprehensive set of resources was created by the authors and uploaded to the module Moodle page to support students in using STACK effectively. These included a reference guide, screencast demonstrations showing examples of how to write code for rendering mathematical equations, and live tutorial sessions where students could ask questions and seek clarification. Additionally, interactive support mechanisms were integrated to ensure students had sufficient guidance throughout the learning process. This reflects the modelling stage of cognitive apprenticeship theory (Knaut et al., 2022). Fig. 1. Illustration of writing basic syntax code for rendering mathematical equations with immediate validation. This Moodle-integrated tool was utilised for both formative and summative assessments to enhance student learning and engagement. The first formative assessment was a syntax quiz, designed to guide students through a large series of exercises where they wrote syntax to render mathematical equations. This setup allowed for multiple attempts, fostering a trial-and-error approach with instant feedback, enabling students to identify and correct errors in real time. The key benefit was that students received immediate validation of their responses, reinforcing their understanding via coaching (see Figure 1). The second formative assessment was a mock exam, structured to simulate the summative assessment without a time limit. Unlike the syntax quiz, the instant feedback was disabled, encouraging students to complete the exam under realistic conditions while still providing an opportunity for self-assessment. This aligns with the fading method of the CAT model outlined by Collins et al. (1991). This practice exam was valuable preparation before the final assessment. The final summative assessment was a time-limited exam, performed in an invigilated university setting, where students applied their learning in a high-stakes environment. While feedback was not provided during the exam, students were still warned if they entered invalid syntax. Once they entered valid syntax, they could view the rendered output as equations, allowing them to self-assess the accuracy of their formatting and notation. Upon completing the assessment, students were given a survey to evaluate the effectiveness of this online assessment tool used throughout the module. They rated the usability and accessibility of STACK in terms of ease of system navigation, using syntax and the clarity of the submission process. Additionally, students were asked about their engagement with preparatory resources such as tutorials, the syntax quiz and mock exam, offering a comprehensive understanding of their learning experience and potential areas for improvement. 3 RESULTS AND DISCUSSION In this section, we first evaluate the effectiveness of STACK as an online assessment tool. Based on survey responses from a cohort of 51 students, valuable insights were gathered regarding their experience with the platform. The results indicate that 89.21% of respondents felt they received sufficient advice and guidance regarding the assessment and module. Additionally, 88.89% agreed that the resources were wellorganised and functioned smoothly, and 87.57% reported that their skills developed throughout the module, highlighting the tool’s positive impact on learning and engagement. Many students highlighted the benefits of the available resources, particularly the syntax quiz and mock exam, which provided valuable opportunities for practice. Positive feedback included comments such as, "I like that there are resources to practice," "Everything was well-structured," "There was a good use of online tools and collaboration," and "I received enough teaching to confidently carry out the assessment." However, some students expressed a desire for additional mock examstyle questions to further reinforce their understanding, with one stating, "I would like to have more mock exam-style questions to practice using STACK syntax." Another student noted a challenge with the platform, commenting, "Worrying about syntax makes it harder to focus on the actual questions." Next, we examine the correlation between students' final exam scores and their engagement with preparatory resources for a cohort of 51 students (see Figure 2). The distribution of student performance categories is as follows: Of the 17 students who achieved a distinction in the final exam, all 17 engaged with the syntax quiz, and 7 also completed the mock exam. Among the 15 students who attained an upper second-class (2:1) final exam’s grade, 14 engaged with the syntax quiz and 5 with the mock exam. For the 9 students who received a lower second-class (2:2) final exam result, all 9 completed the syntax quiz, while 4 attempted the mock exam. Finally, out of the 10 students who scored below a lower second class (2:2), 8 engaged with the syntax quiz, and only 2 attempted the mock exam.. Although the survey included a few self-reported questions on student engagement with preparatory resources, to ensure accuracy, actual engagement data was extracted from Moodle analytics. This dataset, generated from Moodle spreadsheets, provided insights into students' interactions with key resources, including engagement with the syntax quiz and participation in the mock exam. By leveraging this objective data, we aim to assess how resource utilisation correlates with academic performance. Fig. 2. Student engagement with syntax quiz and mock exam across grade categories Figure 2 indicates that students who achieved distinction and upper second-class marks demonstrated higher levels of interaction with both the syntax quiz and mock exam. This emphasises the important role of these resources in supporting academic success. It is noted that even students in lower classifications engaged with these tools in some way, suggesting they were accessible and well-structured. This highlights the importance of providing interactive and inclusive resources that build confidence and proficiency in using the Moodle-integrated assessment tool before the final exam. However, it is important to acknowledge that other factors may have influenced these results. For instance, students with stronger prior knowledge or a greater aptitude for independent learning may have adapted to the system more easily, contributing to their higher performance. While these variables are relevant, this study focuses specifically on the correlation between engagement with preparatory resources and assessment outcomes. Therefore, external factors such as individual learning abilities are not accounted for in this analysis. These insights suggest that while STACK provided structured and effective support, further refinements, such as expanded practice opportunities and additional scaffolding for syntax learning, could enhance student confidence and ensure the tool fully supports their engagement with assessment content. In addition, questions must be carefully designed for STACK to ensure compatibility with its syntax-based notation and automated validation. Proper structuring prevents technical issues, reduces cognitive load, and allows students to focus on problem-solving rather than platform navigation. Well-designed questions improve grading accuracy, provide meaningful feedback, and enhance the overall learning experience by making assessments more accessible and effective. 4 CONCLUSIONS This study contributes to engineering education research and practice by demonstrating how innovative online assessment tools, such as STACK, can enhance student engagement, self-directed learning and assessment effectiveness in STEM disciplines. Stack has potential to address a gap in the coaching stage of the CAT model in online assessment, fostering self-regulated learning through structured feedback mechanisms. The findings align with broader educational frameworks advocating for active learning, constructivist approaches, and adaptive assessment strategies, reinforcing the importance of aligning digital tools with effective pedagogical principles. The broader impact of this research extends beyond this specific implementation. It supports wider efforts to embed digital literacy, inclusive assessment, and automated feedback systems into STEM curricula. These findings contribute to ongoing discussions about scalable and sustainable online assessment practices. This offers insights that can inform institutional policy and future innovations in digital education for engineering disciplines. REFERENCES Ahmed, A. M., & Seid, A. M. (2023). Enhancing Students’ Linear Algebra I Learning Using Assessment Through STACK. International journal of emerging technologies in learning, 18(19), 4-12. https://doi.org/10.3991/ijet.v18i19.42371 Bach, S. (2020). Using multiple choice questions in STACK - reasons and examples https://doi.org/https://doi.org/10.5281/zenodo.3945809 Collins, A., Brown, S.J. & Holum, A. (1991). Cognitive apprenticeship: making thinking visible. American Educator, winter Cameron, R. (2022). Computer Adaptive Practice for a Foundation Mathematics Course. International journal of emerging technologies in learning, 17(23), 2837. https://doi.org/10.3991/ijet.v17i23.35917 Crisp, G., Guàrdia, L., & Hillier, M. (2016). Using e-Assessment to enhance student learning and evidence learning outcomes. International Journal of Educational Technology in Higher Education, 13(1), 1. https://doi.org/10.1186/s41239-0160020-3 Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P. (2014). Active learning increases student performance in science, engineering, and mathematics. Proceedings of the National Academy of Sciences - PNAS, 111(23), 8410-8415. https://doi.org/10.1073/pnas.1319030111 Johnson, E., Weber, K., Fukawa-Connelly, T. P., Mahmoudian, H., & Carbone, L. (2025). Collaborating with mathematicians to use active learning in university mathematics courses: the importance of attending to mathematicians’ obligations. Educational studies in mathematics. https://doi.org/10.1007/s10649-024-10381-x Juma, Z. O., Ayere, M. A., Oyengo, M. O., & Osang, G. (2022). Evaluating Engagement and Learning Based on a Student Categorization using STACK, Exam Data, Key Informant Interviews, and Focus Group Discussions. International journal of emerging technologies in learning, 17(23), 103-114. https://doi.org/10.3991/ijet.v17i23.36619 Knaut, J., Altieri, M., Bach, S., Strobl, I., & Dechant, K. (2022). A Theory-Based Approach of Feedback in STACK-Based Moodle Quizzes Taking into Account Self-Regulation and Different Proficiency of Learners. International journal of emerging technologies in learning, 17(23), 38-55. https://doi.org/10.3991/ijet.v17i23.36425 Kraska, M. (2022). Meclib: Dynamic and Interactive Figures in STACK Questions Made Easy. International journal of emerging technologies in learning, 17(23), 15-27. https://doi.org/10.3991/ijet.v17i23.36501 Lima, R. M., Andersson, P. H., & Saalman, E. (2017). Active Learning in Engineering Education: a (re)introduction. European journal of engineering education, 42(1), 1-4. https://doi.org/10.1080/03043797.2016.1254161 Mejeh, M., Sarbach, L., & Hascher, T. (2024). Effects of adaptive feedback through adigital tool – a mixed-methods study on the course of self-regulatedlearning. Education and Information Technologies, 29(14), 1-43. https://doi.org/10.1007/s10639-024-12510-8 Moradi, M., Liu, L., Luchies, C., Patterson, M. M., & Darban, B. (2018). Enhancing Teaching-Learning Effectiveness by Creating Online Interactive Instructional Modules for Fundamental Concepts of Physics and Mathematics. Education Sciences, 8(3). https://doi.org/https://doi.org/10.3390/educsci8030109 Paiva, R. C., Ferreira, M. S., Mendes, A. G., & Eusébio, A. M. J. (2015). Interactive and Multimedia Contents Associated with a System for Computer-Aided Assessment. Journal of educational computing research, 52(2), 224-256. https://doi.org/10.1177/0735633115571305 Saadati, F., Rohani Ahmad, T., Ahmad Fauzi Mohd, A., & Kamariah Abu, B. (2015). Effect of Internet-Based Cognitive Apprenticeship Model (i-CAM) on Statistics Learning among Postgraduate Students. PLoS One, 10(7). https://doi.org/https://doi.org/10.1371/journal.pone.0129938 Shi, W., & Shahbazi, Z. (2024). Assessment in Online Mathematics and Statistics Courses During COVID-19: Insights from Students and Teaching Staff at a Canadian University. Canadian Journal of Science, Mathematics and Technology Education, 24(1), 57-72. https://doi.org/10.1007/s42330-02400322-4 Shih-Che, L., & Hung-Hsu, T. (2022). Perceived Effectiveness of Developing a Mobile System of Formative Test with Handwriting Revision to Devise an Instruction Design Based on Cognitive Apprenticeship Theory. Sustainability, 14(4), 2272. https://doi.org/https://doi.org/10.3390/su14042272 Sukackė, V., Aida Olivia Pereira de Carvalho, G., Ellinger, D., Vânia, C., Petronienė, S., Gaižiūnienė, L., Blanch, S., Marbà-Tallada, A., & Brose, A. (2022). Towards Active Evidence-Based Learning in Engineering Education: A Systematic Literature Review of PBL, PjBL, and CBL. Sustainability, 14(21), 13955. https://doi.org/https://doi.org/10.3390/su142113955 Ting, F. S. T., Shroff, R. H., Lam, W. H., Garcia, R. C. C., Chan, C. L., Tsang, W. K., & Ezeamuzie, N. O. (2023). A Meta-analysis of Studies on the Effects of Active Learning on Asian Students' Performance in Science, Technology, Engineering and Mathematics (STEM) Subjects. The Asia-Pacific education researcher, 32(3), 379-400. https://doi.org/10.1007/s40299-022-00661-6