Applying Learning Analytics for the Early Prediction of Students' Academic Performance in Blended Learning
The results show that learners’ engagement with the content, exercises and mentoring of a blended learning environment, final academic performance could be predicted when only one-third of the course had elapsed. Furthermore, the blended data set combining online and traditional critical factors (including face-to-face mentoring after class) resulted in the best learner performance.
Journal of Educational Technology & Society, Vol. 21, No. 2 (April 2018), pp. 220-232