Graphing and Interpreting Data

Your firm is located in a 10-story building. Each floor has its own Xerox machine in a
copy room. The firm owns these machines but must pay for paper, toner, and occasional
maintenance. Each employee has a key that opens the copy room door on his or her
floor only and does not have access to Xerox machines on other floors.
Because the Xerox machines are "free goods" right now, you suspect that the firm's
Xerox costs could be cut pretty drastically. To test this theory, you have decided to perform
an experiment. Each person on the 10th floor has been given a card that operates
the 10th-floor machine. These employees have been told that their card will generate a
daily accounting of their Xerox activity. Tenth-floor employees have also been told that
they will not be charged for their use of the machine, but they certainly know that someone
will have some sense of individual usage patterns.
To establish a basis of comparison, the group on the third floor has not been converted
to the card system. The third-floor machine has an internal mechanism that totals
the number of copies made each day, but you do not know who is doing what, and the
third-floor employees have no reason to believe that they are being monitored. The 3rd
and 10th floors have been chosen for this experiment because these two floors have had
about the same usage rates in the past.
You have collected data from the two machines over the last 50 working days. The
data are on the accompanying CD in the file named XEROX. This file has 50 observations
and three variables:
766 CASES
Variable Label
DAY An indicator of the day data were extracted.
TENTH The number of Xerox copies made on the 10th floor.
THIRD The number of copies made on the 3rd floor.

1. Use appropriate graphical or statistical techniques to determine whether the card
accounting system will effectively lower inappropriate Xerox usage if implemented
companywide.

keywords: z-tests, levels, significance