Uluslararası Öğrenci Değerlendirme Programı 2015 Verilerinin Veri Madenciliğinde Kümeleme Yöntemleriyle İncelenmesi

Abstract

In this study, it is aimed to investigate the results obtained from the methods based on PISA 2015 data with the help of Self-Organizing Map, K-Means and Two-Stage Clustering Method. For this purpose, it is aimed to determine how many clusters of students are divided according to different methods, how each cluster is defined and the variables that are effective in these clusters. In the scope of the study, systematic sampling was applied to the students of OECD member countries and as a result analyzes were conducted on 9870 students. The number of input variables used in the study was determined to be five, namely the average of factor scores for four science teaching sub-dimensions and the average of ten plausible values in science . As a result of the study, the ideal number of clusters determined by Self-Organizing Map and K-Means Methods is four and the ideal number of clusters determined by two-stage clustering method was determined to be two. It was determined that different variables were effective in the formation of clusters with the most successful students within the scope of two methods. As a result of the study, it was found that the results obtained from Self-Organizing Map and K-Means Methods were similar in general. As a result of the study, it is recommended that researchers report the results obtained by different methods in clustering analysis. At the same time, clustering analysis was proposed by using R program because of rich results.