Digital Education
Designing scalable, asynchronous, globally distributed postgraduate learning programmes - grounded in evidence, built for accessibility and equity.
UCL STEaPP Online MSc
The STEaPP Online MSc is UCL's fully online, globally distributed postgraduate programme in Science, Technology, Engineering and Public Policy. Launched September/October 2025 - asynchronous, Moodle-delivered, designed for working professionals across time zones with no synchronous attendance requirement.
The programme targeted 20 learners and enrolled almost 40. Uniquely for a fully asynchronous online MSc, no withdrawals were recorded in the first monitored cohort (2025–26) - a significant outcome reflecting the care taken in the programme's design and ongoing support.
My role: structural and pedagogical transition. Working closely with academic staff to move 12 modules from face-to-face and blended formats into fully asynchronous online delivery - building the learning environment, module templates, engagement frameworks and analytics infrastructure.
The Working Environment
Programme development for the STEaPP Online MSc took place in UCL's Learning Innovation Hub - a dedicated space for developing digital and blended learning materials. Building 12 modules for the programme's September 2025 launch required close collaboration with academic staff, iterative review of course content, and regular technical testing across the Moodle environment.
Orchestrating Engagement
An Inclusive Model for High-Growth Online Policy Education
A research paper and conference poster co-authored with Dr Maria Cross (Lecturer and Professional Education Coordinator, UCL STEaPP). Presented at the CODE / RIDE Annual Conference, University of London, March 2026.
The research asked: how do you build a truly scalable, asynchronous, globally distributed postgraduate programme - one that fosters engagement, belonging and academic integrity without relying on synchronous contact or campus presence?
The Design Model
The model centres on a weekly Stimulus → Process → Response cycle: a scenario or case provides the stimulus, structured activities (reading, reflection, analysis) form the process, and peer contributions, forums and structured tasks form the response. The design is consistent across all modules, with a templated weekly format that supports learner orientation and reduces cognitive load.
100%
understood content purpose96.7%
agreed format was suitable90.9%
found examples diverse and relevant44%
spent longer on tasks than estimatedResponding to the Data: Time Estimates and Neurodiversity
The finding that 44% of learners spent longer on tasks than estimated was not treated as a problem with the learners - it was treated as a problem with the estimates. Time estimates across all modules were deliberately reviewed and increased. AI-powered reading time tools were used to generate realistic estimates for reading-based tasks. This was particularly important for neurodivergent learners, who should not be made to feel that taking longer than the estimated time constitutes a failure.
Theoretical Foundations
The model draws on Universal Design for Learning, Community of Inquiry, Cognitive Load Theory, and Constructivism. Assessment integrity in an AI context was a central design challenge, addressed through explicitly structured tasks, authentic assessment prompts, peer review, and reflective submissions that cannot be easily outsourced.
Transitioning 12 Modules to Fully Online Delivery
The core challenge was not building new modules from scratch - it was helping academic staff translate existing teaching into an entirely different format: no required live lectures, seminars or synchronous attendance, and no expectation of synchronous participation. That transition required rethinking pedagogy as much as technology.
Pedagogical Approaches Applied
- Social Learning - with learners distributed across time zones and professional contexts, building peer networks required deliberate design. This included consulting a professional engagement contact from a law firm to understand how geographically distributed professionals build meaningful working relationships - and translating those insights into the programme's peer interaction design.
- Flipped Classroom - some modules included optional synchronous drop-in sessions. The online design ensured that the asynchronous materials were self-contained, so that drop-in attendance was enriching rather than essential - and so that learners who could not attend synchronously were not disadvantaged.
- Cognitive Load Theory - consistent weekly structure, clear objectives, chunked content, and gradual introduction of complexity were all designed to reduce extraneous cognitive load and keep the focus on learning rather than navigation.
Supporting Academic Staff
The transition involved close collaboration with module leads - many of whom were experienced lecturers but had not previously designed for fully asynchronous online delivery. Support included: template frameworks for each week, worked examples of asynchronous discussion activities, guidance on how to write scenario-based prompts, and iterative review of draft course content before go-live.
Accessibility as Standard
Every module was built to WCAG 2.2 AA: colour contrast checking, semantic HTML structure, meaningful alt text, captions on all video content, and clear language throughout. Accessibility was treated as a quality standard, not a retrofit - and was part of the brief given to every academic contributor.
Earlier Work: Immersive Scenario-Based Learning
Proof of Concept - IKON Training (2022–24)
Before the UCL online MSc work, at IKON Training I built a proof of concept to demonstrate the value of immersive, scenario-based eLearning for professional training. The brief was to show that Moodle-hosted eLearning could go beyond static resources and replicate real workplace decision-making in a safe, exploratory environment.
The Lone Worker scenario below was developed as that proof of concept - a branching, decision-based H5P activity placed directly within the IKON Academy Moodle platform. It presents learners with a realistic lone-worker situation and asks them to make a series of safety-critical decisions, providing feedback at each stage. The scenario was designed to demonstrate that interactive, immersive content was achievable within the existing Moodle infrastructure without specialist authoring tools or external hosting.
Why scenario-based learning
For professional training in safety-critical or compliance-heavy contexts, scenario-based learning has a significant advantage over reading or video: it asks learners to apply judgement rather than absorb information. The Lone Worker proof of concept was intended to show IKON's training team and clients that a more engaging, decision-oriented approach was possible - and to make the case for further development of this format across the course portfolio.
Technical approach
The scenario was built using H5P's branching scenario format, embedded directly into the IKON Academy Moodle LMS. No external hosting or licensing was required beyond the H5P Moodle plugin already in place. The H5P format allowed learners to navigate between decision points, receive contextual feedback, and return to alternative decision paths - all within the Moodle environment they were already familiar with.
Try the scenario
The original proof-of-concept scenario is embedded below:
Lone Worker - H5P branching scenario built as a proof of concept at IKON Training, demonstrating immersive scenario-based learning within a Moodle LMS environment.
Online Programme Design Principles
- Consistency over novelty - a repeated weekly structure reduces orientation time and lets learners focus on content rather than navigation
- Asynchronous community - structured peer interaction and authentic forum prompts replace synchronous discussion without losing social presence
- Accessible by design - WCAG-compliant course pages, alt text, captions, plain-language instructions, and multiple content formats
- Assessment integrity - tasks designed to be situational, reflective and context-specific, reducing the risk of AI substitution
- Evaluation built in - Moodle completion tracking, survey checkpoints, and the STEaPP Engagement Tracker provide continuous evidence of design effectiveness
Tools & Platforms
Moodle LMS (UCL instance) Moodle course templates HTML5 / SCSS Panopto (video) SurveyMonkey Power BI (analytics) Dataverse ADDIE Universal Design for Learning Community of Inquiry Cognitive Load Theory Kirkpatrick-Phillips evaluation
Related work
The engagement monitoring infrastructure for this programme is documented in the UCL STEaPP Moodle Engagement Tracker case study. The AI tools and workshops that support both staff and learner AI literacy are documented on the AI for Learning & Assessment page.