Using eye-tracking technology to identify learning styles: Behaviour patterns and identification accuracy

Abstract

Learning style theories have been widely used in adaptive learning systems to enhance learning outcomes. However, the previous studies on adaptive learning systems set a high entry barrier for researchers who lack programming skills, and few of the studies involved authentic everyday learning materials. This author proposes to test the feasibility of eye-tracking technology in identifying learning styles with everyday materials, as well as the identification accuracy. This author selected the Felder-Silverman’s learning style model (FSLSM) as the framework, enlisted the behaviour patterns that can be used to identify the eight learning styles in the FSLSM model, and conducted a quasi-experiment to test whether these behaviour patterns apply to eye movement differences. Then, this author compared the results of eye-tracking identification with participants’ self-report based on Index of Learning Style (ILS) questionnaire for identification accuracy. This quasi-experiment recruited 30 university students, including 19 female and 11 male. Findings showed that eye-tracking technology has the potential to quickly identify learners of different types categorised by the FSLSM theory, with accuracy ranging from 63.50% to 84.67%; however, there are disturbing factors contributing to different levels of identification accuracy, which should be investigated in future research.