Investigation of the predictive power of family and school characteristics for pisa reading skills, mathematics and science literacy by years

This study aims to investigate whether there are variables which significantly predict 15 year old students' reading, mathematics and science literacy by 2003, 2006 and 2009 PISA administrations. It also aims to find out what the variables significantly predicting reading skills, mathematics and science literacy are if there is any. The population of this research of which data were gathered in accordance with predictive correlational research method of correlational researches is composed of all the students who were between 15 years and 3 months and 16 years and 2 months old at the time of PISA administrations. The sample of this research is composed of 4855, 4942 and 4996 students who were selected by the use of stratified sampling method consecutively for PISA 2003, 2006 and 2009 administrations. With this study, by analyzing the Turkish students' responses to common items in the student and school questionnaires administered to them within the scope of PISA 2003, 2006 and 2009 administrations, the correlation between family, school characteristics and the students' reading, mathematics and science literacy was investigated via the multiple linear regression. The data were analyzed through SPSS 20.0 program. Results indicated the family characteristics which were included in the study were better predictors predicting students' achievement than school characteristics. Moreover, it was found out that out of the family-related variables,