Database Reference
In-Depth Information
ing a piece of 2D information into the proper perspective, by means of motion
tracking, really aids comprehension. Hancock and Carpendale [40] consider the
same problem for horizontal displays looking for non-intrusive interactive solu-
tions. Since a study by Wigdor et al.[105] has indicated that angle of viewing
affects readability of certain visual variables; this issue will be an important one
for collaborative information visualizations. This research on how view-angle
distortion affects perception in a single and multi-display environment suggests
that certain types of representations may need to be modified in order to be used
on a digital tabletop display and that information visualizations should not be
compared across multiple display orientations. However, as visual variables were
tested in isolation (e. g. length, direction, only) further evaluations have to be
conducted to see whether participants will correct for possible distortion if the
variables are presented in conjunction with others or whether view correction
[66,40] might compensate.
Visualizations that can be read from multiple angles and orientations (e. g.
circular tree layouts vs. top-down layouts) might be more appropriate for display
on a horizontal surface. However, it is not clear whether participants would try to
read oriented visualizations upside down and make wrong conclusions based on
these readings or whether they would simply re-orient the visualization to correct
the lay-out. Observations of collaborative information visualization scenarios
point in the latter direction [68].
Gutwin and Greenberg [39] discuss issues about viewing a representation in
relation to distributed scenarios. However, parallels can be drawn to co-located
scenarios in which collaborators work with multiple linked copies of the same
representation of a dataset. These essentially represent multiple movable “split
viewports” [39]. The suggested solutions for distributed settings include radar
views, overview+detail solutions, and cursor eye-view. Whether the benefit of
these solutions in a co-located setting outweighs the possible distractions they
might create, however, would have to be evaluated. Further guidelines for using
multiple views in information visualization can be found in [5] and provide a
starting point for tailoring multiple views for collaborative visualization.
Issues of view may develop in collaborative information visualization settings
if collaborators want to switch from loosely-coupled to closely-coupled workstyles
[90] and share discoveries they have made with the other group members. If one
collaborator worked with different view of the dataset it might be dicult to
locate the information in the other persons' view. Another important factor to
consider when developing a collaborative viewing strategy is the establishment
of territories [81] for personal, group and storage purposes that is suggested as
beneficial for group coordination (see “Coordination” above),
A study by Yost and North [109] compared the ability of visualizations to
display large amounts of data normalized across either a small or a large high-
resolution vertical display. Their study showed that the visualizations used were
perceptually scalable but that people preferred different visualizations on the
large vs. small display, as some were found to be easier to read than others de-
pending on the screen size. How these preferences would change for collaborative
work would have to be evaluated.
Search WWH ::
Custom Search