Developing an Automated Essay Scorer with Feedback (AESF) for Malaysian University English Test (MUET) : A Design-based Research Approach

• Main Authors: Bong, Chih How, Ng, Sing Yii, Hong, Kian Sam, Lee, Nung Kion
• Format: Article
• Language: English
• Published: UPM Press 2019
• Subjects: QA76 Computer software
• Online Access: http://ir.unimas.my/id/eprint/30684/1/Developing%20an%20Automated%20Essay%20Scorer%20with%20Feedback%20%28AESF%29%20for%20Malaysian%20University%20English%20Test%20%28MUET%29%20-%20Copy.pdf; http://ir.unimas.my/id/eprint/30684/; http://www.pertanika.upm.edu.my/JSSH.php
• Institution: Universiti Malaysia Sarawak

This paper presents the development of an automated essay scoring mechanism based on the Malaysian University English Test essay marking criteria using the Design-based research (DBR). It is a learning intervention to facilitate students in their essay writing process and at the same time, serves as a tool for teachers to mark essay. DBR is the most commonly used method for conducting research in technological enhanced learning context especially for solving real classroom problem. The development of the automated scoring system is presented step by step following the four phases in DBR model. In each phase, data collection procedure, research instrument and the lessons learnt that lead to further iterations are discussed in order to produce a workable and effective automated essay grader. The outcome resulted from the five iterations lead to the present intervention, Automated Essay Scorer with Feedback (AESF). This system allows teachers to collect samples of marked essays to be trained to grade newly entered essays. Then the teacher can set task and keep track of students’ progress and provide additional feedback as well as rectify the scores generated. For students, they can practice writing essays and demand for feedback at any point of their essays writing process for the system to provide scores by paragraph as well as the whole essay. The system was tested by 24 teachers from 5 schools in real-classroom context with favorable comment.