Automatic classification of incorrect answers to differentiation questions using Potential Response Tree

In this study, the authors focus on Potential Response Tree (PRT) with which users may define rules to classify answers and give feedbacks/partial credit by how they are incorrect or incomplete. Various incorrect answers need to be considered for each question, depending on the combination of elementary functions that make up the question. We categorize differentiation questions and propose an algorithm to classify incorrect answers automatically using a PRT according to the type of questions.