Imagine the following task: you have arranged to meet someone on the university campus at a specific place and on a specific time, or the variant where you have just agreed to meet but haven’t decided on a time yet. Your challenge is to arrive at the right place together with your meeting partner. But something can go wrong: your partner may be late, or went to a wrong location.
For this task, Dearman, Inkpen and Truong created a (wizard-of-oz) prototype for a map application to be used on your location-aware mobile device (such as your smartphone). They instructed participants to perform a number of rendezvous tasks, observed their behavior and noted their comments. Only one person in each pair was a real participant, the other was a fellow researcher following a detailed script.
Although all people were familiar with the campus, only a few people chose not to use the application. The others used it to find their own location, plan their routes to the target location and find out where their partner was. In the first part of the study participants had to manually zoom or pan the map to change the map view, and many of the interactions with the map consisted of preventing yourself, or your partner, to walk off the map. In the second part of the study the prototype was adapted to support this with autofocus and autozoom
For the male participants this automatic zooming and panning seemed to do the trick pretty well. When they were obliged to use the autofocus feature, apparently most of the time they didn’t feel the need to further manipulate the map a lot. Their number of interactions in this condition was significantly different from the female participants: 9.8 for the males (sd 16.9) and 52 (sd 31.6) for the females.
The autofocus as implemented was not ideal though. It did not always show the appropriate level of detail (which made it hard to see in which direction one was moving; and street names were not always visible); landmarks that people would like to use for orientation sometimes fell outside the map boundaries; and the map did not take direction into account (so a participant could be shown at the edge of the map).
Still, the Wizard of Oz approach with a minimal prototype was sufficient to carry out this type of user research with interesting results. Automation of the most frequent user interactions seemed to work well, combined with giving the user the opportunity to temporarily override the automated view, or to switch the autofocus off entirely. Mobile maps show promise for use in social tasks, such as a rendezvous. The map task in this study was different from traditional navigation tasks, because the navigation target was not stationary (the other person is moving too). Other social tasks are even more complex, involving more than two people, or unknown strangers (privacy issues come to mind). When mobile devices become even more common, we’re bound to discover more applications to support such social tasks.
Dearman, D., Inkpen, K., & Truong, K. (2008). Mobile map interactions during a rendezvous: exploring the implications of automation Personal and Ubiquitous Computing, 14 (1), 1-13 DOI: 10.1007/s00779-008-0195-2
