The bank manager in Example 14.3 doubts that the exponential distribution provides a good approximation to the actual interarrival times and service times. Therefore, he collects data on successive...

The bank manager in Example 14.3 doubts that the exponential distribution provides a good approximation to the actual interarrival times and service times. Therefore, he collects data on successive interarrival times and service times on 127 consecutive customers. He then calculates the means and standard deviations of these, with the results shown in rows 5 and 6 of Figure 14.15. (See the Data sheet of the file GGs Template.xlsx.) Are these data consistent with exponential interarrival times and service times? If not, how much do summary measures such as W_Q
and L_Q
change if we use the Allen–Cunneen approximation instead of the M/M/s model? We again assume that there are 6 tellers at the bank.

Objective To see how an approximation to the general multiple-server model can be implemented, and to see how sensitive steady-state measures are to the forms of the interarrival and service time distributions.