Towards Integrated Digital Ecosystems in Higher Education : A Mixed-Methods Evaluation of Two Unified Campus App Concepts

Abstract

Higher education institutions increasingly depend on multiple digital platforms: learning management systems, study administration tools, examination services, resource booking portals, and campus information systems. Although each system exists perfectly independently with its own specific function, their coexistence creates fragmented user journeys, high navigational effort, and disjointed task flows for students. This fragmentation is now occurring at a time when universities are also exploring advanced artificial intelligence (AI) technologies as part of broader digital transformation agendas. However, while the educational benefits of AI are well documented, its capacity to address systemic fragmentation and streamline digital interactions in higher education applications has received little empirical attention.

This thesis addresses these gaps through a three-phase investigation. The first phase comprised an expert heuristic evaluation of existing university systems using a custom 18-point framework tailored to integration-oriented usability. The analysis revealed pervasive issues: disconnected design systems, inconsistent terminology, lack of cross-system linkage, and unclear task flows that collectively increase cognitive load.

The findings established concrete design requirements for unified, or all-apps-in-one, digital ecosystems, leading to second phase, where they were translated into two prototype concepts for a unified campus application. The first is a conventional structured interface designed with clearer information architecture and persistent navigation. The second is an AI-enhanced, conversational interface that enables students to perform academic tasks using natural-language commands. Together, the prototypes showcase two contrasting integration strategies: one with established interaction patterns, and one exploring AI-mediated task execution.

Finally, the third phase involved a mixed-methods usability evaluation with students, combining task metrics, System Usability Scale (SUS) and Raw-Task Load Index (RTLX) ratings, observational data, and post-test interviews. Both prototypes supported complete task success yet produced distinct experiential profiles. The conversational interface yielded fewer errors, no backtracking, and more streamlined task execution, aligning with participant accounts of reduced effort and greater efficiency. The conventional interface, however, was perceived as more predictable, trustworthy, and easier to explore. These qualities contributed to slightly higher usability ratings and more consistent user confidence.

The findings demonstrate that conversational AI can reduce cognitive effort for goal-directed academic tasks, while conventional interfaces remain essential for transparency, exploration, and trust. This suggests that unified university apps should integrate both modes into a “hybrid model” which combines a structured, customizable experience with an optional AI assistant.

The thesis contributes an assessment framework of existing campus systems, two empirically informed prototype designs, and their mixed-methods evaluation. It offers empirical evidence, design guidelines, and methodological insights for universities seeking to develop integrated digital ecosystems and AI-supported student services in higher education.