A Localised Problem-Based Tutor for O/L Mathematics in Sri Lanka

Abstract

This research project will implement a problem-based intelligent tutoring system for the Mathematics curriculum in the General Certificate of Education – Ordinary Level (GCE O/L) government examination, which is taken by all students in Sri Lanka at the age of 16. This system targets improving the performance of students in answering question papers. It will be implemented in the Sinhala language, the main local language in the country.

Every year, the pass rate in O/L Mathematics is around 50%, which has raised grave concern. It is very common in Sri Lanka to seek private tuition especially for Mathematics. However, students from low-income families can only attend large tuition classes and are at a disadvantage when compared with affluent students who can receive individual tuition. Due to the large class sizes at schools and tuition classes, teachers cannot pay individual attention to students, nor do they have time to assess question answers by individual students in a regular basis. The maximum a teacher could do is to simply discuss the questions with the class. Therefore, an affordable problem-based intelligent tutoring system would be very beneficial for students preparing for O/L Mathematics examination. In addition, given that the government has provided computer facilities to schools in the country, such a system is viable and can be accessed by almost all the students in the country in the near future.

Existing intelligent tutoring systems for Mathematics mainly focus on the curriculum in developed countries whereas the Sri Lankan O/L Mathematics syllabus has been developed indigenously. Using a system intended for a different curriculum and examination style would be confusing to students and leave gaps in their preparation. Moreover, in those tutoring systems, feedback is given in English, while most Sri Lankan students have very low English proficiency. Although most of these intelligent tutoring systems can be accessed for free, they are not open source. Thus they cannot be modified for the local O/L syllabus. In this context, a new problem-based intelligent tutoring system in Sinhala for the O/L Mathematics curriculum in Sri Lanka is a timely necessity.

The output of the project will be a web-based intelligent tutoring system for the questions in the O/L Mathematics paper. It will also include a question bank of model questions similar to those asked at the exam, and those given in Grade 10 and 11 term tests. Students will use this system to answer past and model question papers. The system will provide step-by-step feedback on the student’s work in the Sinhala language. As far as we are aware, there is no such intelligent tutoring system for Mathematics that supports any Indic language. Based on the student’s performance on different question types, a model of the student’s knowledge is updated, which in turn would be used to help the student improve.

The main intended user group is O/L students. They can use this tutoring system for free at their convenience. However, teachers can also use this system to conduct more continuous assessments, as they will be freed from the burden of marking papers.

Principal Investigator: Gihan Dias, University of Moratuwa, Sri Lanka

Co-investigator: Surangika Ranathunga, University of Moratuwa, Sri Lanka

Research Site: Colombo, Sri Lanka