Mining learning preferences in web-based instruction: Holists vs. Serialists

Abstract

Web-based instruction programs are used by learners with diverse knowledge, skills and needs. These differences determine their preferences for the design of Web-based instruction programs and ultimately influence learners' success in using them. Cognitive style has been found to significantly affect learners' preferences of web-based instruction programs. However, the majority of previous studies focus on Field Dependence/Independence. Pask's Holist/Serialist dimension has conceptual links with Field Dependence/Independence but it is left mostly unstudied. Therefore, this study focuses on identifying how this dimension of cognitive style affects learner preferences of Web-based instruction programs. A data mining approach is used to illustrate the difference in preferences between Holists and Serialists. The findings show that there are clear differences in regard to content presentation and navigation support. A set of design features were then produced to help designers incorporate cognitive styles into the development of Web-based instruction programs to ensure that they can accommodate learners' different preferences.