SOLVING GEOSPATIAL TASKS IN 2D AND 3D VISUALIZATIONS OF MAPS - Information Visualisation: Techniques, Usability and Evaluation - page 233

n=15. Hence, the number of possible distances is theoretically 120. While
the human visual system is capable of rapidly identifying the 3-5 closest
neighbours (proximity is one of the Gestalt laws of perceptual organization
[20, p. 189]), the difficult part of the task is then rather to correctly
identify the pair with the shortest distance ( D 1 in Fig. 12.3) among the
remaining candidate pairs (e.g. D 2 in Fig. 12.3). The ratio RD=D 1 /D 2 can
be taken as a measure of the degree of difficulty of a stimulus, whereby D 2
is the second shortest distance among all possible distances. With RD=1.0 ,
both shortest distances are equally long, and it is hardly or not possible to
tell a shortest pair of candidates. RD=0.5 identifies a fairly easy to solve
stimulus, as the second shortest distance is already twice as large as the
shortest.
Experimental design
The experiment followed a within-subject design, whereby participants
repeated the experiment three times under different conditions. In those
three trials, the visualisation condition was always chosen to be strong 3D
(S3D), but tilt angles for the map were increased gradually from 45 ͼ to 55 ͼ
and 65 ͼ with respect to the vertical plane (see Fig. 12.4). Tilt angle was a
factor for subsequent statistical analysis. Altogether, participants viewed a
set of 25 stimuli in each of the three repeated trials. Hence, in order to
counteract potential learning effects between subsequent trials, three
different sets of 25 stimulus images, with comparable overall levels of
difficulty, were used.
Test procedure and apparatus
Participants in this experiment were final year bachelor students and staff
at the University of Gävle aged between 22 and 65 years. They were 4
female and 14 male subjects, and all had normal or corrected normal
vision. The majority of subjects had some previous experience with 3D
computer applications, but none of them had tried stereoscopic 3D
displays with head-tracked dynamic perspective before.
Subjects were instructed to identify in each stimulus the two labelled
map positions with the shortest distance to one another. They were told
that accuracy is more important than speed, and there was no time limit to
solve the task. Responses were registered using a conventional mouse
device. The mouse was also used to manoeuvre a 3D crosshair cursor in
the 3D stereoscopic viewing condition. The cursor movement was
constrained to the 2D plane defined by the tilted map which provided
intuitive control.