Adaptive learning model in support of learning the Malay language amongst primary school students with dyslexia

Adaptive learning model acts as an important supplement to the traditional teaching method that commonly uses pen, paper, card or any materials for students with dyslexia. Students with dyslexia have poor fluency in reading, writing and spelling. Additionally, the students also faced difficulties related to engagement behaviour such as giving attention and engaged to the learning materials which constantly need teacher’s intervention. The adaptive learning model or ALMo-DML is intended for students with dyslexia in primary school to learn the Malay language. The model is aimed to address dyslexia cognitive difficulties and engagement behaviour pertaining active participation. Existing learning model covers a limited cognitive difficulty such as confusion, spelling or reading and failed to incorporate adaptation with engagement behaviour. ALMo-DML however, focus on a wider aspect of cognitive difficulties that include phonology, spelling, reading and writing. Cognitive difficulties are identified from the number of errors made in the phonology, spelling, reading and writing exercise given. Additionally, the ALMo-DML incorporate the engagement behaviour through engagement prediction from image classification using machine learning approach particularly Support Vector Machine (SVM). A prototype of the ALMo- DML named as Disleksia Belajar Cerdas Belajar (DBCB) is developed based on the proposed learning model. The work includes (i) cognitive identification using number of errors, (ii) the engagement behaviour prediction developed using frontal face image classification and (iii) the adaptation intervention established using mastery level adjustment, hints and feedbacks. Lastly, the ALMo-DML is evaluated using expert validation for the correctness of the adaptation and experiment for the effect of the adaptation. The expert validation result shows 89% agree that the ALMo-DML was correct in term of the suitability of the adaptation towards the students with dyslexia. Meanwhile, a quasi-experiment is conducted to the 18 students age 7 to 12 years old in Dyslexia Association Malaysia (DAM). The DBCB is compared with MyLexic, a computer-based application to learn the Malay language used in DAM, to see the significant effect on the engagement to the students with dyslexia. The result shows a longer on-task time with an average of 22 minutes when using the DBCB as compared to the MyLexic with only 12 minutes (p value <0.05) using Mann-Whitney U test. As a conclusion, the ALMo-DML able to support the learning of the students through adaptation of cognitive and engagement behaviour with minimal teacher intervention. Despite of that, a few limitations occurred in this research that includes inability to acquire more participants during preliminary study as well as during evaluation. This resulted inadequacy of data relates with student’s image for machine learning prediction. Finally, limitation in scope of DBCB application that only cover basic syllabus of Malay language. Therefore, for further work, the participants can be acquired not only from DAM but also public school. In addition, it also suggested for future work to cover a wider area of syllabus in Malay language especially writing skill that currently still lack in research.

Text FSKTM 2020 16 IR.pdf Download (1MB)

Actions (login required)

View Item