EXPLORATION BY EXPERIMENT
As a foray into intermodal interface design, an experiment was devised based on the activities that take place in a typical EOC.*
In this experiment, subjects were given audio messages emulating those received by a typical emergency communications center. The messages described the occurrence of various emergency events (e.g., robberies and automobile accidents) ' within a selection of neighborhoods, and the resulting sets of activities completed by the assigned emergency services resources. The messages were scripted to report each stage of each resource (e.g., police, fire department, and emergency medical services) working through its assigned event (e.g., enroute to the incident, arriving at the scene, and returning to stand by). The audio was prerendered by a text-to-speech software utility utilizing a high-resolution American English voice model. The timing between messages was controlled to provide a slow period of activity (messages arriving approximately every 10 seconds), which would then ramp up and sustain a period of fast activity (messages arriving approximately every 4 seconds). Subjects were selected to use one of three graphical user interface (GUI) variants with which to capture the activities reported to them in the audio messages. The GUI variants all had the same basic arrangement of screen elements representing neighborhoods, resources, incidents, and so on, but had some differences that were the focus of the experiment.
Interface #1 was text-only (though still utilizing GUI point-and-click input). Interface #2 had the same screen layout as interface #1, but had parts of the screen shaded in different colors corresponding to the different types of emergency service: police, fire, and medical.* The third interface variant was identical to interface #2, with the addition of a text-output area where transcripts of the audio messages would appear as the messages were heard. This was effectively a simulation of a high- accuracy and high-speed voice recognition system (Figures 16.1 through 16.3).
FIGURE 16.1 Screen-shot of interface #1 (no status data displayed).
Any form of color blindness or other uncorrected problems with vision were disqualifications from participation, as was any problem with hearing or any experience with emergency services dispatching.
FIGURE 16.2 Screen-shot of interface #2 (no status data displayed).
FIGURE 16.3 Screen-shot of interface #3 (no status data displayed).
-  * “Emergency Operations Centers” (EOC’s) and “Emergency Communications Centers” (ECC’s) aresynonymous for the purposes of this work. Both are often referred to as “9-1-1 centers” in the UnitedStates.
-  A set of seven different incident types were specified, each type requiring a unique combination ofemergency resource types (police, fire, and medical). For example, a robbery required only a policeresponse, whereas an automobile accident required all three services to respond.