Towards adaptive open learning environments: Evaluating the precision of identifying learning styles by tracking learners’ behaviours

Open learning represents a new form of online learning where courses are provided freely online for large numbers of learners. MOOCs are examples of this form of learning. The authors see an opportunity for personalising open learning environments by adapting to learners’ learning styles and providing adaptive support to meet individual learner needs and preferences. Identifying learning styles of learners in open learning environments is crucial to providing adaptive support. Learning styles refer to the manner in which learners receive and perceive information. In the literature, a number of learning style models have been proposed. The Felder and Silverman Learning Styles Model (FSLSM) has been selected as the most appropriate model for open learning. In previous studies two approaches have been used to automatically identify learning styles based on the FSLSM. These approaches are known as the data-driven method and the literature-based method. In the literature, the literature-based method has been shown to be more accurate in identifying learning styles. This method relies on tracking learners’ interactions with the provided learning objects based on a set of pre-determined patterns that help in inferring learning styles. The patterns are monitored based on pre-identified threshold values. This paper aims to apply the literature-based method to open learning environments and introduce the optimal patterns and threshold values for identifying learning styles based on the FSLSM. To achieve this aim, a study was conducted whereby a prototype that simulates the open learning environment was developed and piloted on an undergraduate IT course so that learner behaviour could be tracked and data could be collected. Next, different sets of threshold values from the literature were considered along with some updated threshold values considering the context of open learning environments, and the precision of identifying learning styles was calculated. Eighty-three students participated in the study and used the developed prototype. Precision results from different threshold values presented in the literature along with customised threshold values for this study are reported and analysed in this paper. It is shown that threshold values derived from literature and customised to suit open learning environments provide a high level of accuracy in identifying learning styles. The paper presents the first study of its kind in evaluating threshold values and precision in identifying learning styles based on the FSLSM in open learning environments. The results are promising and indicate that the proposed methodology is efficient in detecting learning styles in open learning environments and useful for developing an adaptive framework.