Systematic Literature Review: An Investigation Towards Finding Constructs For Performance Prediction of Students in an Online Engineering Course

Abstract

The use of technology in the field of engineering education has been the most common intervention, especially in the post-pandemic era. Most universities have plans to continue the blended approach to education in the future. This decision has to be an evaluated decision and the analysis of a student's performance serves as an input to take the decision. The other advantage of analysis include early prediction of student performance which can help the instructors to provide timely interventions and help the students to improve their performance. Thus identification of constructs that reflect student engagement and performance in a course delivered in online mode is very essential. This literature review attempts to bring forward the constructs used by various researchers that reflect student engagement and performance. The review is situated in the context of engineering education delivered in online mode. The identification of constructs is significant and helps to build machine learning models for predicting the performance of the students. Standard Systematic Literature Review(SLR) methods defined in literature including citation searching and hand searching were carried out to identify the constructs that have been in use. A list of constructs used by researchers in the literature, that capture students' attention and performance are identified and presented in this review. The identified constructs include students' interaction with content, students' interaction with peers, demographic factors, and the academic records of the student. These validated constructs are proposed to integrate with the Learning Management System(LMS) and use the feature for early prediction of student failures.