Intelligent Tutoring Systems for Software Engineering Education in Nigerian Higher Institutions: A Design Theory

The persistent gap between software engineering education outputs and industry expectations in Nigeria calls for innovative pedagogical technologies that provide personalised, adaptive learning support at scale. This dissertation develops a design theory for intelligent tutoring systems (ITS) in software engineering education at Nigerian higher institutions, contributing original design theory to educational technology and software engineering education research. A design science research methodology guides three artefact development cycles. Phase one reviews ITS design theory literature and software engineering education pedagogy research, establishing theoretical foundations. Phase two conducts a contextualisation study through 25 interviews with software engineering lecturers and 60 student surveys at three Nigerian universities spanning South East (FUTO), North Central (University of Jos), and South South (University of Benin). Phase three designs, implements, and evaluates a prototype ITS for introductory software engineering concepts, incorporating Bayesian knowledge tracing for learner modelling, misconception-triggered explanation generation using large language models, and adaptive practice problem sequencing. The prototype is evaluated with 80 undergraduate students in a randomised controlled trial over 12 weeks. Students in the ITS group achieve a 22 percent higher post-test score gain than the control group (p = 0.004), with the effect significantly stronger among students from low-bandwidth access zones. A Nigerian ITS Design Theory is formulated, comprising 12 design principles and 8 meta-requirements grounded in Nigerian higher education constraints. Keywords: intelligent tutoring systems, software engineering education, design theory, Nigeria, adaptive learning