This study attempted to identify the information and communication technology items that affected students' mathematics and science literacy scores by making use of the 2015 PISA data, The presence of numerous items related to ICT in the PISA and the administration of these items to large groups of people provides researchers with a large data source. However, researchers experience challenges in revealing the significant and beneficial data among the entire data set. So one of the most commonly used data mining method is the Chi-squared Automatic Interaction Detection method (CHAID), which is the decision tree method. As a result of the CHAID analysis, conducted to reveal the ICT items related to mathematics literacy scores, it was revealed that there was a significant relationship between mathematics literacy scores and the eight variables. For science literacy, there was a ten significant relationship variables. There is a relationship between high science and mathematics literacy scores and using digital devices at an early age as well as feeling comfortable with using digital devices at home. As an outcome of the CHAID algorithm, the realization of a significant reduction was achieved in the dimensionality of both models. The selected variables can be used for future research and development of new, parametric models. In the resulting model, apart from the reduction of the number of predictors, the reduction of their categories was also achieved.