The determination of learner groups with cluster analysis method in open and distance learning systems

The aim of this research is to separate the learners registered in the Open Education System of Anadolu University in 2014-2015 academic year into clusters by analyzing their features and to lead the determination of strategies appropriate for the types and features of learners. Cluster analysis, which is one of the descriptive data mining methods, was used to obtain learning groups in this study which is using the obtained data set by combining the questionnaire data applied to 43,106 learners studying in the Open Education System and the data of these learners in the institution database. With the cluster analysis, experimental studies were conducted on the data in IBM SPSS Modeler 14.2 in order to obtain the groups and the algorithm suitable to the features of the data set was specified as TwoStep. As a result of the research, three groups were obtained with the TwoStep algorithm made on the prepared data set. The ratio of these clusters ranges from 36.9%, 35.7% to 27.5%. The most distinctive features in the determination of the clusters were specified as academic final grade, monthly income, working sector, working hours, sex, internet access facility, marital status, media ownership, and age. It is thought that the results of this study will facilitate taking decisions for learner-based academic, administrative, corporate and support services, and support the differentiation and design of the services according to the features of these groups. Keywords: Open and Distance Learning Systems, Grouping e-Learners, Cluster Analysis, Segmentation, TwoStep Algorithm.