Investigation of Change in Reading-Math-Science Literacy Scores in PISA Applications with Multivariate-Multilevel Model
Abstract
The aim of this study is to determine the variables of school and student level that significantly affect the math-reading-science performance of Turkish students participating in PISA 2009-2012-2015 applications. Within the scope of this aim, gender, type of program, educational status of parents, educational opportunities at home, educational resources at home, educational facilities at home, economic-social-cultural status index were used. Regarding school level, school size, computer usability index, teacher-student ratio were included in the model. For the purpose of the study, data of PISA 2009-2012-2015 Turkish students were analyzed using the Multilevel-Multivariate Regression Model. According to the findings, it was observed that there were differences in math-reading-science scores among the schools. PISA 2009-2012 Turkey application of students' math-reading and science scores together predictor as revealing the student level, respectively, parental education status of variables, home educational resources, and in PISA 2015 application maternal education level, it is determined that the economic-social-cultural situation is. In all three applications, it was found that gender was the variable that predicted mathematics and reading performance significantly. PISA 2009 school-level predictors of Turkey as meaningful to the students of math-reading and science scores along with the application variables were found to be the teacher-student ratio. For PISA 2012 and PISA 2015, it was found that the school level variable which predicted academic achievement significantly was the usability index of computers. Based on the findings of the research, it may be suggested to reduce the disparity in student achievement among schools. Studies can be done to regulate the students' educational environments at home.
Keywords: PISA, multivariate-multilevel model, academic achievement.
DOI: 10.7176/JSTR/5-11-10
To list your conference here. Please contact the administrator of this platform.
ISSN (online) 2422-8702