Classification of Students' Mathematics Learning Achievement on Bloom's Taxonomy-Based Serious Game Using Ordinal Logistic Regression

Abstract

Students' profiles classification is needed to make learning more focused on the purpose of learning. Student profiles that related to student cognitive abilities are generally seen from their learning achievement. Student learning achievement is obtained through assessment instruments. One alternative assessment that can used to measure students' mathematics learning achievement is Bloom's Taxonomy Based Serious Game (BoTySeGa). BoTySeGa's output are consists of three attributes that can be used as material for classifying students' profile. Three attributes are classified into student learning achievement categories of insufficient, sufficient and good. Classification is carried out using ordinal logistic regression method, where the results called as classification predictions and compared with the actual classification value that is obtained from students' mathematics learning achievement tests. Level of accuracy classification between prediction classification results and the actual classification results is obtained by 55% in moderate category.