Understanding a student’s academic ability to pass the national examination for pharmacists with the pass/fail prediction model

Abstract

This study used the pass/fail prediction derived from the Support Vector Machine (SVM) to understand the academic ability of individual students to pass the national examination for pharmacists. In a previous report, we confirmed the accuracy of the pass/fail prediction model using scores from a mock examination given in September. In addition to this exam, the pass/fail prediction model was applied to the mock exams conducted in November and January. With the pass/fail prediction as an index, the progress of individual academic ability of the students was investigated throughout the term. It was found that students with high academic ability in September were more likely to pass the national examination than those who gradually improved over the course of study after September. Also, the progress in learning status for individual students was difficult to grasp when using the total score of the mock exams but was easier to understand when using pass/fail prediction as an index.