Cloudworks is no longer accepting new user registrations, and will be closing down on 24th June 2019. We hope to make a read-only archive of the site available soon after.
Learning analytics and Predictive modelling
Cloud created by:
2 October 2018
You are warmly invited to the next eLearning Community event on Learning analytics and Predictive modelling.
- 10.00-10.30 OU Analyse - The models - how it works? What's under the bonnet? Future development direction - Martin Hlosta (Research Associate, STEM/KMi)
Abstract - OUAnalyse (http://analyse.kmi.open.ac.uk) is a project piloting machine-learning based methods for early identification of students at risk of failing. We primarily focus on identifying students who are at risk of not submitting (or failing) the next Tutor Marked Assignment (TMA). This has been identified as a proxy for the possible future student failure and also a close milestone to focus in case an intervention is needed. The predictions are generated weekly and the students with their risk of failure are available in the dashboard to the Course tutors and Module Chairs to consider appropriate support. The predictions and other supportive information within the dashboard support the course users in better understanding of their students. In my talk, I will describe how we generate the predictions, which data we use and also describe our current work on the Personalised Study Recommender.
10.30 - 10.45 Coffee
- 10.45-11.15 Student Probabilities: Predictive Model for Supporting Students - Galina Naydenova (Senior Modeller; Planning, Forecasting and Modelling Strategy & Information Office) and Claire Maguire (Student Experience Manager, U/Grad & Prof Dev)
Abstract - The Student Probabilities model produces probabilities, or predictions, of an individual student reaching specific milestones in their module (for example, completing the module, returning in the next academic year). These predictions are based on models built from observed behaviour of students in previous academic years, applied to the current cohort. This information can be used in various ways - for example, in helping to identify students to target for interventions; in identifying modules where the pattern of success differs from the norm; in managing student cohorts on modules of particular interest. All of these uses could result in increasing student success. We have evidence that early interventions carried out by SSTs using the Student Probabilities for selection have beneficial impact on student retention.
The presentation will cover the Student Probabilities model from data/modelling point of view.
11.15-11.45 The EAI project - evidence generated so far and future direction - Christothea Herodotou (Lecturer in Innovating Pedagogy, LTI, IET) and Avinash Boroowa (Senior Product Development Manager, LTI, TEL)
Abstract - OU Analyse (OUA) is a system that predicts students at risk of not submitting their next assignment. It has been used in more than 48 modules across the university during the last three academic years (2015-2018). This presentation presents evidence about the effectiveness of OUA for supporting teachers and students through a systematic mixed-methods evaluation (interviews, focus-groups, statistical comparisons, learning analytics data). Evidence suggest that OUA use by teachers is a significant predictor of students’ performance and retention, with average teachers’ engagement leading to better learning outcomes, over and above teachers’ and students’ general performance. Qualitative data point to positive views of involved stakeholder in using predictive data widely across the university, with teachers in particular recognising its usefulness in complementing the teaching practice and monitoring students’ behaviour online.
You are welcome to attend for all or part of the event, and we will aim to stick to timings to enable you to do that. If you are unable to attend in person, please join us via the live stream on Stadium on Wednesday 10 October 10.00-12.00.
This event will also be recorded and can be viewed afterwards via the Stadium link above.
If you are planning to attend in person please email firstname.lastname@example.org so we can order enough refreshments. Do contact us if you have any questions. Please also pass on details of this event to any of your colleagues who might be interested in attending.
OU staff can access presentation slides after the event at: Scholarship Exchange