Feedback for Faculty on Student’s Asynchronous Learning Based on Classification using Topic Modelling

Abstract

The emphasis on using virtual and, Information and Communication Tools (ICT) is increasing in higher educational institutions to enhance the general education standards for 21st century students. As a part of this, blended learning is emerging as a significant way for course delivery, especially after the hard times of COVID-19. One of the challenges of blended learning is inadequate information for faculty to understand how each student in their class has engaged with the content of the asynchronous video. Reading, assessing, and grading the long and reflective assessments is also tiresome. Therefore, in this study, we aim to provide feedback for faculty by classifying students into different levels based on the reflective answers’ content after the asynchronous learning mode. The site and context of the study are one of the modules named "Sustainability in Engineering" in the course "Engineering Exploration," offered to first-year undergraduate students. We adopted the qualitative text mining approach called Latent Dirichlet Allocation (LDA) widely used topic modeling technique to understand if the topics discussed by students align with the asynchronous content. The results of the study show that 12% of students have spoken only on sustainability, 25% of students have spoken about only engineering design, 68% of students have tried to establish a relationship between the sustainability and engineering design concept, and 20% of students spoke none of the above three topics.