In user-controlled CIDCW, the subscriber heard a 440-Hz tone for 0.3 s, while the far-end party received a silent break for 0.3 s.
After both conditions ended, the subjects directly compared automatic and user-controlled CIDCW. The assessments were made using seven-point scales with labeled end points.
From the figure it is clear that both forms of CIDCW received excellent ratings, whether the subjects were considering that form of the service alone or comparing it with Call Waiting service.
After completing both conditions, the subjects were asked to make direct comparisons of automatic and user-controlled CIDCW. Figure 3 lists the questions that were asked and the mean ratings.
Automatic CIDCW was rated as being easier to use than user-controlled CIDCW, and it was preferred over user-controlled CIDCW.
When the subjects were directly asked which form of CIDCW they preferred, 18 selected automatic and 6 selected user-controlled.
Full use of the user-controlled form of CIDCW requires that subscribers initiate two breaks into their conversations - one to find out the name and telephone number of the new caller and the second to talk to the new caller.
It should be noted that this strategy might not be a practical way to handle the interruptions if user-controlled CIDCW were actually deployed.
It was observed that the subscribers made two different types of errors in using CIDCW, both of which occurred in the user-controlled condition.
In Call Waiting service and in automatic CIDCW, the one action subscribers can take (flashing the switch hook) switches them between the far-end party and the new caller.
With the increasing use of voice mail, then, it would be more likely with user-controlled CIDCW than with automatic that new calls subscribers wanted to take would be forwarded to voice mail before they could be answered.
After completing both conditions, subjects were asked to make direct comparisons of automatic with silence and user-controlled CIDCW. The mean ratings are shown in Figure 5 (left graph).