Okuma Becerilerini Yordayan Özelliklerin Belirlenmesi: Genetik Algoritma Kestirimi

Abstract

The current study aimed to determine the features predicting students’ reading skills. The study group of the study was comprised of 5232 students aged 15 years from 42 countries participating in the PISA 2015 application. The data of the study were obtained over the PISA 2015 program and analyzed by using the regression model based on the genetic algorithms method estimation. It was intended to perform a feature selection process for the regression model, which consisted of the variables that best predicted reading skills with the method of genetic algorithms. The results obtained revealed that the variables of gender, father’s education level, use of the internet at home, the language used at home, the number of e-book readers, the speed at which the items measuring the reading skill was responded and the number and variety of the books at home significantly predicted the students’ reading skills. The variation in the variables whose prediction power was found to be significant was understood to lead to a significant variation in