The 2-D, 3-D, and 3-D
(augmented with vertical bar graph) displays were user
tested for accuracy and performance. Results from this
analysis revealed that user response times were decreased
by 23% with a reduction in errors of 60% using the
standard 3-D display. Additional testing is needed to
determine the benefit of the vertical bar graph.
Keywords
Graphical User Interface, Real-time, Air Space Monitoring
Introduction
Real-time Graphics User Interfaces (GUIs) involve unique
challenges for display designers. Real-time displays for
the DoD and the Department of Transportation (DoT) are
being investigated for monitoring air space using 3-D
formats. Previous research has shown that although most
users prefer 3-D data presentations over 2-D, there can be
detrimental effects in user performance[3]; however,
monitoring the position of missiles and airplanes can best
be visualized using a 3-D coordinate system. This
experiment examines the possibility of enhancing user
performance using a 3-D environment. The standard 3-D
environment was also augmented using a vertical bar graph
display concept. The results of this study are significant
since user performance for a DoD or DoT display can
impact a number of lives. The experiment described in this
report addresses the area of 3-D formats for real-time
displays. The following questions were assessed:
- Although users almost always prefer 3-D data
presentation over 2-D, people do not always correctly
judge what features of an interface will provide them with
the best performance. Consequently, they may express
preference for interfaces that have degraded
performance[4]. Will using a 3-D display environment
enhance performance for specified air space monitoring
environments?
- Will augmenting the 3-D display with an integrated
vertical bar graph display showing information such as
signal strength improve user performance?
Methodology of Experiment
The experiment was designed to measure user
performance with regards to time response and accuracy of
actions. The user was assigned the following tasks for
each format test:
- Discern the specified enemy targets flying in space
(red conical objects)
- Select the enemy targets using the mouse
(more than one enemy target may be selected)
- Engage enemy targets by selecting a button
(represents launching a weapon)
If the user accidentally selects objects other than enemy
targets, he or she could deselect the objects by clicking the
mouse on them a second time. Engagement of any objects
other than enemy targets (balloons, heavys, and unknowns)
were considered illegal and recorded as user errors. The
three format tests were configured using the same number
of objects in random locations and motions. All subjects
were tested for each of the three formats in the same order
(2-D, 3-D, and 3-D with vertical bars).
Format one displayed objects in a 2-D format similar to
currently used real-world interfaces. The operator has
options to show error ellipses, and quickly perceives via
the color-coding and shapes, which objects require
immediate attention. The operator also has the option to
remove individual objects of other classification upon
selection. This format gives users very little ability to
select objects that were partially or fully hidden behind
other objects.
Format two displays objects very similarly to the 2-D
format; however, the 3-D capability allows users to rotate
the space on two different axes using scroll bars to obtain
different perspectives. A vertical scroll bar rotates the
entire 3-D space forward or backward on an axis which
runs horizontally to the users view on the screen. The
horizontal scroll bar rotates the 3-D space around an axis
which runs vertically to the user's view on the screen. The
vertical axis might not be displayed vertically on the screen
but may be inclined at some angle off vertical.
Format three also displays objects in 3-D space but
incorporates vertical select bars. The selectable vertical
bars utilize the fact that the human eye scans faster
vertically than horizontally[1,2]. This format takes
advantage of this physiological fact by placing each
vertical bar under its respective object on the 3-D display.
The vertical bars represent a value such as signal strength
of the object or other measurable quantity. The color-
coded bars are displayed geographically directly beneath
the position of the object on the screen which provides an
enhancement for the user to quickly correlate the value
represented by the vertical bar with its associated object.
When the user moves the horizontal scroll bar to reorient
the 3-D space, the vertical bar graph is also reoriented to
keep each vertical bar displayed directly beneath its
respective object. When selecting an object using this
format, the user may move the cursor to select the object in
3-D space or select the object's respective vertical bar
which automatically selects the object in 3-D space. When
an object is selected, the object is highlighted by a green
octagon and its respective vertical bar is highlighted by a
green rectangle. This capability would aid the operator if
the enemy targets' vertical bars were higher than those of
other type objects. As a result, the operator could just
move the cursor horizontally across the peaks of the
vertical bars and select all the enemy target objects.
FIGURE 1. Augmented 3-D Display with Vertical Bars
Experiment Results
Six different subjects completed each of the three formats
three times for a total of 18 tests per format. This data was
then analyzed using an ANOVA (Analysis of Variance)
test to determine if there was a significant difference
between the formats. The F values revealed that there was
a significant difference between the 2-D and 3-D formats.
Although test results suggest an improvement with
augmenting the 3-D format with integrated vertical bars,
there was no significant statistical difference between the
basic 3-D format and 3-D with vertical bar graph. The
basic 3-D display allowed the user to rotate the field of
view so that closely spaced targets could be engaged. This
display exhibited the best user-performance resulting in a
decrease of user-response times by 23% along with a 60%
reduction in errors. The 3-D display with vertical select
bars had a similar reduction in user-response times when
compared to the 2-D format. It showed promise, for
certain users, of improvement over the basic 3-D format.
Conclusion
These results revealed that the 3-D formats were superior
to the 2-D format considering user-response times. The
apparent advantage of augmenting the 3-D format using
the vertical bars is still undetermined. Testing the three
formats in different order might show interesting results
but was not investigated here. Further testing is required to
determine the benefit of the integrated vertical bars.
Testing of the 2-D format with the vertical bar graph might
also reveal better performance over the basic 2-D;
however, the ability to rotate the air space seems apparent.
Acknowledgements
The research described in this paper was conducted under
the U. S. Army Small Business Innovative Research
Program contract DASG60-92-C-0121. The contracting
agency is U.S. Army Space and Strategic Defense
Command, Huntsville, Alabama.
References
[1] The Defense Intelligence Agency (DIA) Standard User
Interface Style Guide for Compartmented Mode
Workstations (DIA 1989, DDS-2600-6215-89).
[2] Fernandes, K. User Interface Specifications for Navy
Command and Control Systems, Version 1.1. U.S.
Department of the Navy; Naval Command, Control, and
Ocean Surveillance Center, Research, Development, Test,
and Evaluation Division, San Diego, California (1992).
[3] Reising, J.M. and G.L. Calhoun. Color display
formats in the cockpit: Who needs them? Proceedings of
the 26th Annual Meeting of the Human Factors Society
(pp. 446-449), Santa Monica, CA: Human Factors Society
(1982).
[4] C.D. and A.D. Andre. Performance-Preference
Dissociations, SID International Symposium Digest of
Technical Papers, (pp. 369-370), Santa Ana, CA: Society
for Information Display (1994).
