Learning analytics : validating the efficacy of digital simulations in higher education

![CDATA[The Horizon Report (Johnson et al, 2012) provides ‘foresight’ advice that Learning Analytics (LA) is trending to adoption in a time horizon of two to three years. While LA has emerged from work in Educational Data Mining (EDM) emphasising applications to learning over 15 years, the advent of ‘big data’ and predictive analytics drawn from Learning Management Systems (LMS) and other university data sources has led management to expect increasing business intelligence. Termed ‘Academic Analytics’ this business focus is, in part, on student behaviour, success and particularly the effectiveness of university services to students. In contrast the focus of LA is squarely on student learning and should more immediately benefit students, teachers and designers. As Web-based digital simulations (eSims) are increasingly drawn into sophisticated blended learning environments driven by mobile devices and cloud computing, it becomes more possible for eSim instructional design to benefit from various forms of data science. This paper suggests potential benefits that the emergent fields of ‘data visualisation’ and LA may bring to the challenge of delivering validated eSim designs to mainstream higher education. A brief case study of the DeakinSim ‘Forensic Interviewing of a Child’ will clarify several possible contributions that LA may offer our ongoing Design Based Research (DBR) approach at Deakin University. LA can assist DBR to improve and validate the University’s suite of DeakinSims for eLearning, facilitating student learning in real-time, but also help teachers ensure they are producing the specified professional skills in graduates that DeakinSims are designed to address.]]