Highly-informative competence driven feedback for digital learning (HICOF-DL)

Using artificial intelligence and learning analytics technologies, the project team designed and implemented novel ways of providing highly-informative competence driven feedback (hicof) to large classes of university students (≈1000) for three innovative online assignments. We built machine learning models to automatically assess the assignments. High-informative feedback provides correct solutions, possibilities for improvement, hints on self-regulation and effective learning strategies, and on two different levels - Feed Up (concerning their understanding of the learning goal) and Feed Forward (concerning their next steps in their learning process) (Hattie & Timberley, 2007). Competence driven feedback is feedback that is given to students based on a competence framework (set up by the course responsible) consisting of all the learning goals and objectives of the course, and the skills and knowledge that students should gain after course completion. All of the course assignments are associated with a number of learning objectives, and feedback is given based on a calculation of the competences that students showed in their submitted assignments.

A pilot study was completed to collect students’ responses in their assignments in order to build machine learning models, which could be used to automatically assess them in the evaluation study and provide hicof to students. Learning analytics indicators were used to collect information relating to what students have understood. The three assignments are 1) a text-based essay where students should address ten items that were discussed in the (online) lecture and commonly presented in the short online tutorial, 2) a forum discussion assignment where students should collaboratively complete a learning task and 3) a concept map assignment, where students should demonstrate their overall understanding of the topics and subtopics being taught by displaying the branches and nodes in the correct arrangement.