Adapting Learning in Online Courses through Automatic Detection of Learners’ Learning Styles 13 March, 2013
Dr. Kinshuk, Athabasca University, Canada led the weekly discussion on 13th March 2013 with his talk on “Adapting Learning in Online Courses through Automatic Detection of Learners’ Learning Styles”.
Online courses are no longer a luxury for universities. Some universities are using it heavily these days. Again, if classes are of small size, a teacher can check how students are attentive and provide adequate feedback as well. However, in case of massive online courses, this is not very practical.
Dr. Kinshuk talks about a lot of thoughts on e-learning and education technology in Journal of Educational Technology & Society.
There is a lot of research happening in the field of Technology Enhanced Learning (TEL) today. Learning Management Systems (LMS) are often used in TEL. LMS Packages like Moodle, Blackboard, Sakai etc. are mostly developed to support teachers to create courses.
However, it needs to be noted that in spite of technology used in education, teachers cannot be replaced by technology. Teachers provide pedagogy required for the process of education. LMSs provide various features to ensure pedagogy by teachers and interaction among students. They are domain independent and content in one LMS can be reused by other LMSs.
Still, LMSs provide only little or in most cases no Adaptivity. Also, there is no consideration for students in LMSs, since they are teacher centric.
Every student has different needs, characteristics and situations. So, it is important to consider individual differences and provide a personalized learning experience to students. There is a need to focus on different needs and characteristics of students. Several factors such as Prior Knowledge, Motivation to learn, Learning goals and interests, Cognitive abilities and Learning styles guide the individual learner.
There are several adaptive systems such as Aha!, Tangow, Inspire etc. in the market today. Still, they have a lot of limitations as they consider only few needs and characteristics of learners. They are developed for specific content or features and lack support to teachers. Also, content in such systems cannot be reused.
An important consideration here is how individual characteristics and needs of learners can be considered in learning systems to improve learning experience. Several factors such as Learning styles, cognitive differences etc. needs to be considered here. For instance how do students with different characteristics behave in TEL? Again, how to indentify students’ learning characteristics?
There are different contexts of adaptive learning such as LMSs, Ubiquitous learning environments, Collaborative Learning and Game Based Learning.
There are about 70 different learning styles such as Active/ Reflective, Sensing/Intuitive, Verbal/Visual Sequential/Global etc. A lot of research has gone into learning styles in the last 30 years.
There are still a lot of issues over learning styles. There are questions researchers try to answer such as whether learning styles are stable over time, can learning styles be measured etc. Relationships between different learning style models are still not clear.
Felder Silverman Learning Style model claims that each learner has a preference on each of the above dimensions (Active/ Reflective, Sensing/Intuitive, Verbal/Visual Sequential/Global). In fact, the model does not categorically call a learner “Active” or “Verbal”. Instead, it classifies learners in a spectrum (Mostly Active, Balanced between sensing and intuitive etc.). This is done on basis of a 44 item questionnaire (11 questions per dimension).
How to identify Learning Styles
Collaborative Student Modeling:
The questionnaire method of identifying learning styles has some limitations. Students have to take time out of their schedule to fill the questionnaire up. Also, non intentional influences like what their teachers would think might prompt them to fill up wrongly. Furthermore, questionnaire can be filled up only once. It is more or less “a snapshot of learning styles”, which is bound to change over time.
Automatic Student Modeling:
In this method, learning styles are identified based on interactions of students with the course. There is no additional work for students. The method analyses data from a specific time span. It is more accurate and allows tracking of changes in learning styles.
However, the method has some challenges such as getting reliable information to build a robust student model.