Abstract

This paper presents an application and validation of the use of Markov chains for modelling and analysing group performance. The model combines a Markov process with reliability theory. Inter–arrival times of work are treated as “failure times”, while service times are considered as “repair times” of each worker. Each row in the transition matrix is a binomial distribution showing the probability of finishing various amounts of work. The state of the process represents the number of currently occupied workers of the group. Experiments on work groups were conducted to confirm the validity of the model. The experimental results show agreement with the theory. The study indicates that Markov chains can be used to examine: (1) the group’s probability of carrying out work when it arrives, (2) the availability of staff, and (3) the group utilisation. The paper provides a practical tool for managers in modelling and analysing performance of a work group. However, the research is limited to types of work with constant arrival rates and to situations where work is lost if not begun immediately on arrival.