Modeling of students’ cumulative grade-point average (CGPA): A Markov chain approach

This paper focuses on applying some basic concepts in discrete-time Markov chains to the modeling of students’ Cumulative Grade-Point Average (CGPA), using records of their first year in school. The primary focus is on forecasting the proportion of students that transit from one grade to another, as they journey through their academic semesters. The Fundamental matrix is obtained for the absorbing Markov chain, and other indices like the initial distributions, expected sojourn times for the transient states, the n-step transition probabilities and the absorption probabilities are obtained for the process. The results show that state 1(First class) is an absorbing state, which makes its expected sojourn time to be infinity. The expected sojourn times for states 2, 3, 4, 5, 6 have respectively been found to be 6.67, 10.00, 3.45, 1.67, 1.54. the probability of absorption to state 1, from the other (transient) states are all zeros.