The Future of Education
Predicting Drop-Out Rates in E-Learning Courses
Katharina Resch, Die Berater - Bridges to Europe (Austria)
Abstract
Background:
Teachers and tutors with a great amount of teaching experience typically have a lot of information in their heads in order to predict learning outcomes and whether or not a student will fail a class. This intuition – if you could call it like this – is based on analytic competences of the teachers, observation, and experience with different learner’s groups. Actually you could say that they have “data” in their heads. Faculties, institutions, or education providers cannot access or use this information systematically if it only exists in the heads of teachers. Thus, measuring and extracting this information, processing it through a prediction algorithm to predict learning outcomes, does make sense. In the United States this has already become standard practice for schools, universities, and colleges (Dietz-Uhler & Hurn 2013).
Methods:
The phenomenon of dropping out of e-learning is described as a multi-factorial phenomenon in the literature (Simpson 2006), meaning that it needs more than just a few indicators to accurately predict success of failure. The number of indicators varies from 3 to over 40 in different empirical studies, but generally the more indicators the higher the accuracy of prediction.
In the European project CRITON (2014-2015) a prediction system is built, based on an indicator model with 4 main indicator groups
- Socio-demographic indicators of learners
- Individual learning and efficacy indicators
- Behavioral indicators in e-learning
- Management indicators
which shall be presented at the conference.
Results:
The results show the different indicators (40 in total) which are necessary for predicting drop out in e-learning. A practical prediction system for teachers was developed based on these indicators.
Literature:
Dietz-Uhler, B. & Hurn, J. (2013): Using learning analytics to predict (and improve) student success: a faculty perspective. In: Journal of Interactive Online Learning, 12(1), 17-26.
Simpson, O. (2006): Predicting student success in open and distance learning. In: Open Learning 21(2), 125-138.
