From Adoption to Actionability: Investigating Learning Analytics Integration in Learning Design for Higher Education Teaching Practice

While learning analytics has gained significant attention in higher education, educators find it a constant challenge to translate learning analytics insights into actionable learning design decisions. This thesis investigates the integration of learning analytics in learning design through a theoretically grounded approach anchored in self-regulated learning, guided by three research questions: RQ-1: What actionable insights can be drawn from the relationship between learning analytics and learning design? RQ-2: What barriers and enablers do educators face when adopting learning analytics in learning design? RQ-3: What is the relationship between learner personality traits and learner engagement across different learning design activities? RQ-1 examined the practicalities required for successful integration through a systematic literature review combining bibliometric and thematic analysis of 85 empirical studies. The critical finding revealed that only 18.8% of studies demonstrated empirical evidence for actioning learning analytics in learning design, exposing widespread low adoption. This established the theoretical foundation through the Learning Analytics integrated Learning Design framework while imposing the investigation of implementation barriers, laying the foundation for RQ-2. RQ-2 investigated what prevents educators from using learning analytics in learning design, building on RQ-1's low adoption rates. A survey of 120 educators across 34 institutions in 8 countries identified factors affecting adoption patterns across educators in Computer Science and Engineering and other disciplines. These findings constructed the organisational onion framework, capturing technical, formal and informal level factors contributing to adoption. This revealed that systematic attention to multiple organisational layers is required; while educator perceptions on motivating factors highlighted the need to understand individual learner differences for effective integration, providing the foundation for RQ-3. RQ-3 addressed the empirical gap for closing the loop of actioning learning analytics, examining learning behaviours among different personalities across learning design activity types. Analysis of 72 learners across 100 activities categorised according to the Open University Learning Design Initiative taxonomy identified four distinct learner archetypes: resource explorers, consistent achievers, collaborative learners and delayed starters. This provides empirical evidence captured through learning analytics, for personalising learning design based on individual learner characteristics. This research makes distinct contributions towards actionability and adoption of learning analytics-informed learning design. Theoretical contributions include a learning analytics integrated learning design framework, organisational onion model for adoption, and learner personality-based learning design guidelines. Empirical contributions comprise adoption rates across faculties, and learner personality-engagement correlations. Practical contributions propose an assessed toolkit with implementation guidelines. Methodological contributions incorporate a mixed-method approach combining bibliometric analysis, survey methodology, and learning analytics data analysis. This research transforms learning analytics and learning design integration into systematic, evidence-based practice in higher education teaching.

Name dc:type.qualificationname

PhD in Computer Science

Level dc:type.qualificationlevel

PhD thesis

Grantor dc:publisher.institution

Queen Mary University of London

Year dc:date.issued

2026

Author dc:creator

• Sooriya-Arachchi, C.

• Learner Analytics, Learning Design, Self-Regulated Learning, data-driven decision-making

Identifier

oai:westminsterresearch.westminster.ac.uk:x57zv

OAI identifier oai:identifier

oai:westminsterresearch.westminster.ac.uk:x57zv