Question: A retail manager in a discount store wants to establish a policy of the number of cashiers to have on hand and also when to open a new cash register. The first step in this process is to determine the rate at which customers arrive at the cash register. One day, the manager observes the following times (in minutes) between arrivals of customers at the cash registers:
a. What kind of theoretical distribution do you think would be appropriate for these data? Why?
b. Run @RISK's fitting procedure on the data. Is the distribution that you choose in part a the best fitting distribution? Does the distribution you choose in part a fit the data closely enough to be chosen?
c. We did not use all the information we had when we ran the preceding fitting procedure. Waiting times can never be negative. Therefore, rerun @RISK's fitting procedure setting the lower bound to be fixed at zero. See Steps 3.5 and 3.6 for instructions on how to modify the fit. Does this improve the fit of the distribution you choose in part a? How do the mean values compare?