Abstract:

The objective of this study was to identify the variables that could predict the PISA reading comprehension success of countries with different success levels. This was done with 34 independent variables obtained from the student questionnaire given to the students who participated in PISA in 2018. For this purpose, 79 countries participating in PISA were put in order of success percentage. Then, with equal slices, countries were divided into lower, middle, and upper slices. Significant variables in predicting reading comprehension success in the analysis made with three different logistic models based on the data of low, middle and high achievement group countries in percentiles varied between 25-28. Afterwards, the number of variables was reduced by adjusting the classification performance of the models created in such a way that there would not be a notable decrease, and the 10 variables as the best-predicted success were selected for each model. As a result of the analysis made with the models created by reducing the number of variables, it was revealed that the important common predictors for the low, middle and high success group countries such as educational items at home, reading pleasure, socio-economic status index, PISA test's perception of difficulty, family wealth, meaning of life and teacher's orientation to education. Then, with 10 independent variables in each of the models created for each group, the models' performance in classifying and predicting PISA Reading Comprehension success was calculated.

Keywords: