Mining students' learning patterns and performance in Web-based instruction: a cognitive style approach

Abstract

Personalization has been widely used in Web-based instruction (WBI). To deliver effective personalization, there is a need to understand different preferences of each student. Cognitive style has been identified as one of the most pertinent factors that affect students' learning preferences. Therefore, it is essential to investigate how learners with different cognitive styles interact with WBI programs. This paper presents an empirical study, which examines the effects of cognitive styles on students' learning patterns and the effects of learning patterns on their learning performances. Riding's cognitive style analysis was used to identify the students' cognitive styles. Data mining, especially a clustering technique, was used to analyze the results. It was found that field independent students frequently used an alphabetical index whereas field dependent students often chose a hierarchical map. Such learning patterns also have great effects on their performance, especially for field dependent students.