An Analysis of PISA 2018 Mathematics Assessment for Asia-Pacific Countries Using Educational Data Mining

Source abstract

The purpose of this paper is to determine the variables of high importance affecting the mathematics achievement of the students of 12 Asia-Pacific countries participating in the Program for International Student Assessment (PISA) 2018. For this purpose, we used random forest (RF), logistic regression (LR) and support vector machine (SVM) models to classify student achievement in mathematics. The variables affecting the student achievement in mathematics were examined by the feature importance method. We observed that the variables with the highest importance for all of the 12 Asia-Pacific countries we considered are the educational status of the parents, having access to educational resources, age, the time allocated to weekly lessons, and the age of starting kindergarten. Then we applied two different clustering analysis by using the variable importance values and socio-economic variables of these countries. We observed that Korea, Japan and Taipei form one group of Asia-Pacific countries, while Thailand, China, Indonesia, and Malaysia form another meaningful group in both clustering analyses. The results we obtained strongly suggest that there is a quantifiable relationship between the educational attainment and socio-economic levels of these 12 Asia-Pacific countries.