The use of Markov Models as an aid to the evaluation, planning and benchmarking of doctoral programs

In this paper, an absorbing finite Markov chain model is developed to facilitate planning in a DBA Program in a Graduate School in Australia. The short- and long-term forecasting facilities offered by the model are used to determine the expected numbers in the DBA Program. The Markov model is also used to determine other critical attributes of the Program, such as expected success rate for candidates, expected first passage times prior to absorption per se and also prior to withdrawal and graduation specifically, all facilitating a status quo analysis. The absorbing state-specific first passage times are arrived at using a relatively simplified approach. This paper also illustrates how the model can be used to competitively benchmark these status quo attributes. The expected number of doctoral candidates in each state within the Program (in both the short- and long-term contexts) can be used to estimate, among a range of important model attributes, the expected revenue stream and expected supervisor load in the DBA Program. From this information changes can be made to the Program where required.