Machine learning and deep learning models are pivotal in educational contexts, particularly in predicting student success. Despite their widespread application, a significant gap persists in comprehending the factors influencing these models’ predictions, especially in explainability within education. This work addresses this gap by employing nine distinct explanation methods and conducting a comprehensive analysis to explore the correlation between the agreement among these methods in generating explanations and the predictive model’s performance. Applying Spearman’s correlation, our findings reveal a very strong correlation between the model’s performance and the level of agreement observed among the explanation methods.

This paper describes an approach to help students involved in a Programming Tutoring System, providing them with feedback during the coding problem-solving activities. It provides feedback for students during the coding, helping them to fix mistakes and how to take the next steps to complete the solution. This way, the student does not need to complete and submit a solution to get feedback from the system. The approach uses three feedback resources: videos, text hints, and flowcharts. We conducted an experiment involving 34 students from a programming introduction course. Preliminary results indicated a positive impact on the students learning. Our results also suggested that we can provide valuable feedback to students with difficult to complete a solution.