Identifying the Factors Affecting Science and Mathematics Achievement Using Data Mining Methods
Keywords:
Integrated science and mathematics, data mining, TIMSS, PISAAbstract
The purpose of this article is to identify the order of significance of the variables that affect science and mathematics achievement in middle school students. For this aim, the study deals with the relationship between science and math in terms of different angles using the perspectives of multiple causes-single effect and of multiple causes-multiple effects. Furthermore, the study examines and reveals how the reading skills, problem solving skills, cognitive and affective variables influence the math and science achievement. The data was collected from the results of Turkish students who participated in three international examinations; TIMSS 1999, PISA 2003 and PISA 2006. We analyzed the data using two data-mining methods (decision trees and clustering). The findings show that science or mathematics achievement is not influenced by the course-specific variable alone but also by other related variables. The following variables are the most important; the students’ reading and problem-solving skills affected both mathematics and science achievement; the mathematics achievement affected the science achievement; and the science achievement affected the mathematics achievement. It is also found that the affective variables have almost equally significant effects on the science and mathematics achievement.