2026 ASEE Annual Conference & Exposition

Beyond Measurement: Mapping the Design and Theory of Adaptive Formative Systems in STEM and Engineering Education

Presented at Educational Research and Methods Division (ERM) Poster Session

This theory work-in-progress paper investigates the conceptual and methodological development of Adaptive Formative Systems (AFS) in STEM education. Rooted in Black and William’s (1998) notion of assessment for learning, formative assessment is intended to improve learning through actionable feedback rather than merely measure achievement. Yet, many technology-enhanced “adaptive” systems typically built on Item Response Theory (IRT), Computerized Adaptive Testing (CAT), or ELO-based algorithms have optimized measurement precision while neglecting the formative purpose of assessment. As Qadir et al. (2020) assert, formative assessment is “for aiding, not grading.” However, most adaptive tools in STEM contexts still automate grading rather than enable feedback, reflection, and instructional adaptation. This disconnection motivates a systematic re-examination of how adaptivity can authentically enact formative assessment principles to enhance student learning.
Guided by the PRISMA-ScR framework, this review synthesizes literature across Engineering Village (Compendex & Inspec), Scopus, IEEE Xplore, ERIC, and Web of Science databases using search terms such as “adaptive learning,” “adaptive assessment,” “formative feedback,” “assessment for learning,” and “STEM education.” Over 8,000 records were retrieved, reflecting both the growth and conceptual fragmentation of adaptivity research. The study is driven by three research questions:
(1) What are the key design features and adaptive mechanisms that characterize current adaptive formative systems?
(2) What techniques, models, and algorithms (e.g., Bayesian networks, reinforcement learning, cognitive diagnostic models) are used to personalize feedback and assessment pathways?
(3) What learning outcomes and theoretical frameworks underpin these systems, and to what extent do they align with formative assessment theory?
Preliminary analyses show that most adaptive systems emphasize cognitive adaptivity adjusting item difficulty or sequencing while rarely integrating affective or behavioral data to guide motivational feedback or learner support. Few explicitly anchor their design in formative theory or demonstrate how feedback leads to learning improvement. These findings reveal a persistent gap between algorithmic adaptivity and pedagogical intent.
The anticipated outcome of this work is a conceptual and design taxonomy for Adaptive Formative Systems that operationalizes the formative pillars of cognition, observation, and interpretation (Qadir et al., 2020). Insights from this synthesis will inform the next phase of the author’s dissertation: developing and testing a CAB-based (Cognitive, Affective, Behavioral) adaptive formative assessment prototype aimed at transforming assessment from a static measure into a dynamic, feedback-driven process that truly supports learning in engineering education.
Keywords: formative assessment; adaptive systems; feedback; engineering education; learning analytics

Authors

VINCENT OLUWASETO FAKIYESI University of Georgia [biography]

Vincent Oluwaseto Fakiyesi is a doctoral student in the Engineering Education program at the University of Georgia. He holds a bachelor’s degree in Chemical Engineering and is interested in advancing assessment practices to better understand student learning. His research focuses on cognitive diagnostic modeling, student misconceptions, and adaptive assessment systems. He applies diagnostic models to analyze large-scale educational assessment data and generate actionable insights that support teaching, learning, and personalized feedback in engineering education. He also examines assessment practices in the era of generative artificial intelligence, emphasizing adaptive, interpretable, and learning-centered systems.
Dr. Nathaniel Hunsu The University of Georgia [biography]

Nathaniel Hunsu is an Associate Professor of Engineering Education at the University of Georgia, affiliated with the Engineering Education Transformational Institute (EETI) and the School of Electrical and Computer Engineering. He leads the Engineering Student Learning and Engagement Research (ESLER) group, mentoring students in projects that explore how engineering learners engage with tasks, technologies, and learning environments. His research focuses on conceptual understanding, epistemic cognition, and student engagement, with an emphasis on evidence-based strategies that enhance learning in engineering education. He applies quantitative research designs, systematic reviews, meta-analysis, and assessment instrument development to advance both the theory and practice of engineering education. Dr. Hunsu’s work also investigates the use of emerging educational technologies, including AI, XR, and robotics, in fostering active and self-regulated learning. His goal is to bridge research and practice, equipping educators and learners to achieve meaningful, lasting learning outcomes.
Anthony Fakiyesi Faktlens Technologies Limited [biography]

Anthony Fakiyesi is a cybersecurity professional and researcher with a Master’s degree in Cybersecurity and Human Factors from Bournemouth University. His work focuses on the design and analysis of complex digital systems, with particular interest in how data-driven and intelligent systems can be used to support decision-making, learning, and system resilience.

His research interests extend to the intersection of cybersecurity, artificial intelligence, and adaptive systems, particularly in understanding how user behaviour, system feedback, and data analytics can inform the design of responsive and secure digital environments. Drawing from his experience in security monitoring, threat analysis, and human-centred security, he is interested in how adaptive and interpretable systems can be designed to support both technical performance and user understanding.

His broader research direction explores how principles from cybersecurity, such as risk modelling, behavioural analysis, and system feedback loops, can inform the development of adaptive and data-driven systems in domains including education, where personalised feedback and system responsiveness are critical.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 21, 2026, and to all visitors after the conference ends on June 24, 2026

« View session