Modeling a Hierarchical System with Double Absorbing States

Lydia Musiga, John Owino and Patrick Weke.


This paper describes a Markov Chain transition model for predicting expected numbers of successful graduates and unsuccessful dropouts from an education system, thus the Double absorbing states. This is an improvement on the Single Absorbing State model (see Musiga et al 2010) where successful and unsuccessful dropouts were grouped together. The major limitation of the Single Absorbing State model was the inability to determine the proportions of students who successfully graduated from and unsuccessfully dropped out of the system. The theory of Absorbing Markov Chains is used due to its adaptability to the representation of absorbing rates, transition rates and dropout rates for both successful and unsuccessful students, from an education system. The Double Absorbing States model enables us to predict the numbers of expected qualified personnel vis-à-vis the numbers of unsuccessful dropouts from a system.

Full Text [PDF]