Learning Analytics Dashboards (LADs) can provide lecturers and learners with insights about the learner’s progress with their studies through visualisations of the learner, the learner’s cohort and learning data. Despite a LAD’s usefulness to support learning, very few studies have considered how LADs can be used to identify ‘at-risk’ learners based on programme-specific learner ‘personas’, identified using a human-centred design approach. Different learner personas, in the context of this study, refer to similar aspects of learners' character that are presented to or perceived by others. A learner persona can be enriched through demographic, qualification and career information, learning data, engagement data and assessment scores amongst others.

Although there have been several studies on the development of LADs for higher education, these studies have not specifically emphasised the use of a Human-Centred Design (HCD) approach nor the identification and use of various learner personas in the LAD visualisations. The HCD approach may better comprehend the lecturers' and programme coordinator’s requirements and produce more stimulating insights of various ‘learner personas’ that can be developed and tested based on their involvement in designing the LAD.

This study presents the findings of HCD approach to designing a LAD to identify personas for high, average and low-performing learners for a hybrid online structured master programme. The main emphasis was on identifying the low-performing or ‘at risk’ learners so that actions could be taken to reduce the likelihood of the learner failing a module and/or dropping out of the programme. We present an HCD approach involving a design cycle that employs paper and interactive prototypes to guide the systematic and effective design of a LAD that meets the lecturer’s and the programme coordinator's requirements for a hybrid online structured masters programme, by actively engaging them in the design process. We then conducted usability testing to refine the LAD so that it can be used by a lecturer in near real-time to identify the at-risk learner persona for a particular module, for further intervention. Our findings reveal the process used to develop a LAD based on historic and near real-time learner engagement and assessment data to develop an accurate ‘at risk’ learner persona for a module in a hybrid online structured masters programme. The benefits of the study include that ‘as risk’ learners can now easily, ‘proactively’ be identified at any point in the module syllabus using the LAD and based on their learner persona so that interventions can be made to increase learner throughput and success.

Keywords: Learning Analytics Dashboard (LAD), Human-Centred Design (HCD), ‘at risk’ learners, online, structured Masters, learner personas.