Using an Intelligent Tutoring System to Predict Mathematics and English Assessments

Abstract

Intelligent tutoring systems (ITS) supplement traditional learning by providing personalized instruction. Predicting student performance in formative and summative assessments can help educators and parents determine suitable learning interventions. In this article, interaction log data from three south Indian schools using Amrita Learning ITS were gathered and analyzed. We investigated the extent to which information from the system improves the prediction of students' performance on both formative and summative assessments. Results indicated that prediction improves significantly for both formative and summative assessments when compared to models that only use pretest information.