ABSTRACT
The COVID-19 pandemic has changed dramatically the way how universities ensure the continuous and sustainable way of educating students. This paper presents the neural network (NN) modelling and predicting students’ progression in learning through a hybrid pedagogic method. The hybrid pedagogic approach is based on the revised Bloom’s taxonomy in combination with the flipped classroom, asynchronous and cognitive learning approach. To evaluate the effectiveness of the hybrid pedagogic approach and the students’ progression in learning, educational data is collected that comprises of labs and class test scores, as well as students’ total engagement and attendance metrics for the programming module considered. Conventional statistical evaluations are performed to evaluate students’ progression in learning. The NN is further modelled with six input variables, two layers of hidden neurons, and one output layer. Levenberg-Marquardt algorithm is employed as the back propagation training rule. The performance of neural network model is evaluated through the error performance, regression and error histogram. The NN model has achieved a good prediction accuracy along with limitations. Overall, the NN model presents how the hybrid pedagogic method in this case has successfully quantified students’ progression in learning throughout the COVID-19 period. Copyright © 2021 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
ABSTRACT
The education approaches in the higher education have been evolved due to the impact of covid-19 pandemic. The predicting of students’ final performance has become more crucial as various new learning approaches have been adopted in the teaching. This paper proposes a statistical and neural network model to predict students’ final performance based on their learning experiences and assessments as the predictor variables. Students’ learning experiences were obtained through educational data analytic platform on a module that delivered the mixed-mode education strategy using Flipped classroom, asynchronous and cognitive learning in combination with the revised Bloom’s taxonomy. Statistical evaluations including multiple regressions, ANOVA correlations are performed to evaluate the appropriateness of the input variables used for the later Neural Network output prediction. The Levenberg-Marquardt algorithm is employed as the training rule for the Neural Network model. The performance of neural network model is further verified to prevent the overfitting issue. The Neural Network model has achieved a high prediction accuracy justifying the students’ final performance through utilising the aforementioned pedagogical practises along with limitations. © 2022, Springer Nature Switzerland AG.
ABSTRACT
Experimentation is a key component of any engineering education, and it can be either hardware or simulation (software) based experiments or mixed. While the teaching and learning are provided in an alternative manner (mostly online) for ensuring the continuity of education and the student learning experiences due to the COVID-19 pandemic, both hardware and software based laboratory assessments must also be conducted in an alternative manner. In this paper, we share and discuss a systematic approach as an alternative laboratory assessment (ALA) for Multimedia Engineering modules in the Transnational Education (TNE) programme between Queen Mary University of London (QMUL) and Beijing University of Post and Telecommunications (BUPT). We have conducted this study to verify how the ALA using the systematic approach for two multimedia modules have achieved the intended learning outcomes. Based on the quantitative analysis, we have suggested how we can improve the laboratory assessment in the alternative approach.