On the use and estimation of the transition probabilities of Markov processes

Abstract

Steady state conditions have often been imposed on many Markovian systems. In analysing such systems heterogeneous groups are lumped together; for example repeaters in the educational system are treated just like regular students who often pass at first attempt These conditions have serious implications on the use and estimation of transition probabilities in the Markov analysis. This paper examines some special transition probabilities and discusses how and why they must be properly articulated in order that the ensuing analysis lead to acceptable results.