Investigation of Variables Explaining Science Literacy in PISA 2015 Turkey Sample

Abstract

The purpose of this study was to examine the variables that explain science literacy with the answers given by Turkish students to the Program for International Student Assessment (PISA) 2015 student questionnaire. The Turkish sample of the research, which was conducted in a relational scanning model, is composed of 5895 students selected through a stratified sampling design. The sample of the study consists of 3052 people who remained after the data containing missing values were removed. In this study, Chi-squared Automatic Interaction Detection (CHAID) method, which is one of the data mining decision tree algorithms, was used for data analysis. As a result of the analysis, it was concluded that the variable that best explains the science literacy of Turkish students is “the number of books in the home”. Other variables explaining the science literacy of Turkish students were also investigated in detail. In the research, gain values were examined to determine the most effective node in separating successful and unsuccessful students. As a result, it was seen that the most effective node consisted of students with science self-efficacy among students who had more than 200 books at home, and those who had more than 40 course hours in a week at school.