Forecasting educational enrolments: Comparison of a Markov chain and circuitless flow network model

Entrants into large complex educational systems typically follow one of many different paths to a successful graduation, transfer-out or drop-out. In contrast to the usual simple Markov chain simulation model employing ‘stock data’, the educational system is presented as a circuitless flow network model employing sub-population student attributes and ‘flow data’. A simple linear model readily projects flow patterns into future grade enrolments. The statistics of a secondary school cohort are used to illustrate the methodologies and contrast the results of the Markov and circuitless flow models.