Predicting At-Risk Students’ Performance Based on Their Interactions and Assessments in Foundation Year English Courses at King Abdul-Aziz University

EasyChair Preprint no. 6965

Abstract

Predicting online learners’ final course achievement is most of the time dependant on their course interactions with LMS-based graded assessments, which can be used as an indicator to identify at-risk students in various academic contexts; ultimately, support students’ success and academic advancement. This is where Learning Analytics (LA) representing learners’ behavior inside the Learning Management Systems (LMS), and Deep Learning (DL) techniques come into play as academic data, which can be used as the gateway to informed predictions of learners' future performance. Not surprisingly, the need for developing at-risk profiling models becomes apparent in cases when a large student cohort is taking a foundation course online, for example, and where instructors cannot easily or effectively monitor their students’ progress in real-time. The proposed study aims to utilize Artificial Neural Network (CNN, Encoder/Decoder LSTM, and GRU); to develop models that predict students’ final grade (pass or fail) based on their interactions with the assessments’ types (assignment or test). The LMS data used in this study was for 3,929 Students enrolled in a four-module English foundation course in King AbdulAziz University (KAU) which was collected during the second semester of 2020.  The results show that the CNN model with the lowest RMSE value (RMSE = 3.15) performed better than the other models.

Keyphrases: Artificial Neural Network, Educational Data Mining, Learning Management System, Predict at-risk student

@Booklet{EasyChair:6965,
  author = {Amnah Alsulami and Maiada Almasre and Norah Almalki},
  title = {Predicting At-Risk Students’ Performance Based on Their Interactions and Assessments in Foundation Year English Courses at King Abdul-Aziz University},
  howpublished = {EasyChair Preprint no. 6965},

  year = {EasyChair, 2021}}