Exploration of factors predicting the accuracy rate of the English reading section of the College Scholastic Ability Test using the decision tree model.

The study aimed to identify factors that could affect the accuracy rate of the English section of the College Scholastic Ability Test (CSAT). A total of 45 variables related to the passages and questions that influence the rate of accuracy were included. The decision tree regression model was used to find out the possible variables and analyze the impact of each factor. The decision tree model revealed that the question type had the greatest impact on the accuracy rate of the English reading comprehension test, followed by the points of the question, adjective variation (ADJV), the number of dependent clauses per T-unit (DCT), the number of clauses per T-unit (CT), word count (nwords), and topic. In conclusion, this study shows that question-related factors have a greater impact on the accuracy rate than passage-related factors. Furthermore, this study provides implications for teaching and learning related to English reading in the College Scholastic Ability Test.