New Perspectives in Science Education

This poster reports on the use of an automated feedback system used to deliver feedback to large cohorts of first-year undergraduate students. The system uses data from the VLE (Virtual Learning Environment). Engagement was captured through learners' participation in online quizzes (>35,000 attempts) and pre and post-class video viewings (>10,000 viewings). In class participation and attendance was recorded using Moodle attendance and Lecturer marking which is recorded through the VLE. Data was collected and automated feedback forms based on learners' unique grades and levels of participation within specific learning activities were generated. Learners received the written feedback form from lecturers during high attrition periods of the academic year. Key to creating the personalised feedback forms is a set of algorithms which can be readily modified to ensure the feedback given to students is relevant and focused on their own learning pathway, while not requiring the lecturer to manually assess each student. The process was carried out in three stages. Criteria were selected for student behavior and combinations of behaviors across learning activities. Responses for each of these results or behaviors were created. Finally, the feedback was allocated using algorithms for delivery to students. Survey responses from students indicated that they enjoyed the high level of active engagement, expressing satisfaction with the feedback cycle implementation. 95% of participants asked about feedback (n= 229) indicated that it was “Useful” or “Frequent and Useful”. This presentation examines the evolution of the feedback system over the previous three years as it has changed to meet students needs. This presentation intends to inform practice and will provide a framework for to harness student data and provide personalised feedback to large groups without need for additional IT requirements.

Keywords: Learning Analytics, Feedback, Automation.