Generating grade baseline marks for forcasting students´ performance in Lower Secondary Assessment (PMR)

The present grading system used in schools does not provide a realistic projection of the students´ performance during public examinations such as the Lower Secondary Assessment (PMR). This is because the actual PMR grades obtained by the candidates are usually significantly better than the school trial examination grades using the guidelines provided by the Ministry of Education. A set of minimum marks for all grades, known as baseline data is used to forecast the student´s grade for the forthcoming public examination based on the student´s trial examination marks. To date, such method to determine the baseline data is done manually and is subjective as it depends on a teacher´s personal viewpoint. This paper presents an automated system that generates subject baseline data using empirical data comprising actual trial examination marks with their corresponding PMR grades for a cohort of candidates. We present our findings using data obtained at PMR level.