Don't walk like an Egyptian : Coping with shared attention in a mobile 3D system

Abstract

This thesis presents our study of a system using real-time three-dimensional graphics on handheld computers focusing on the impact of the shared attention problem. The main objective is to investigate whether a mobile RT3D system can be both usable and useful. We postulate that one of the greater challenges to such a system is the problem of shared attention. The dynamic context of truly mobile IT use means that the user will have to share his attention between operating the system and relating to the world around him.

In order to investigate this issue we developed a prototype of a Mobile 3D system. We relied on literature studies, interviews with experts and other research of relevance. As a result we formulated a set of system requirements with the intention to cope with the shared attention problem. We then performed an experiment to test how the mobile 3D system developed was utilized and how our design choices affected shared attention among the test subjects.

The resulting system was received favourably by the test subjects. We observed the subjects obtaining a dynamic pattern of use where they located the destination of each task prior to movement, then checked while moving that they were on the right track. The attention demand of the system was not observed to be intrusive as the subjects appeared to be able to make efficient use of the system while moving. Although the system was received favourably by the test subjects, we suspect that our initial suppositions need to be revised. Our finds indicate that our focus on shared attention may have blinded us to other important factors. There were indications that the subjects had constructed a mental map of the geographical area prior to beginning movement and so did not use the system to aqcuire new information while they were walking. This interpretation leads to several needed revisions of our shared attention model.

In our discussion we argue that focusing our design on shared attention to such an extent is not necessarily ideal for M3D systems. However the start/stop paradigm of the subjects indicates the need for a dynamic interaction design: aimed at rapidly switching between high and low attention modes.