THE DISCRIMINATION OF LEARNING STYLES BY BAYES-BASED STATISTICS: AN EXTENDED STUDY ON ILS SYSTEM

Yanping Jing, Bo Li, Na Chen, Xiaofeng Li, Jie Hu, and Feng Zhu

Keywords

Bayesbased statistics, index of learning styles, data mining, machinelearning

Abstract

Educational data mining (DM) is an emerging discipline, concerned with developing methods for exploring the unique types of data that come from the educational context. As one of the methods, Index of Learning Styles (ILS) system is well designed to identify personalized learning style. However, a remarkable discrepancy between the results of ILS system and participants’ self-estimation indicates a relatively constrained applicable range of ILS in learning style identification. In this study, we focused on working out data-mining methods to extend applicable range of ILS, which is achieved by constructing a new questionnaire system and applying novel DM methods to a group of participants. According to our analysis, Bayes-based statistics are found to be effective in distinguishing ILS classes, and a newly constructed classification system – tree map – can help to distinguish learning style for samples from ILS “neutral class. Therefore, the DM technique applied in this study can be an effective method for enlarging the applicable range of traditional ILS system.

Important Links:



Go Back