Geography Reference
In-Depth Information
Empirical Investigation of Sound Map Usability
Existing Studies
Based on the aforementioned analytical review it is still necessary to perform user
studies aiming at an optimization of different representation methods and respective
parameter settings. There are quite a few examples from other disciplines like
scientific visualization (Bly
1982
) or statistics (Mezrich et al.
1984
); however, an
explicit and profound application on cartographical representations has not been
performed to our knowledge.
A small user study that picked out selected aspects for the representation of
quantitative data for a single point of time was described in an earlier publication of
our research group (Schiewe and Weninger
2012
). General aim of this study was to
assess the overall usability through a comparison of acoustic versus conventional
graphical coding methods. Each test person received four maps that used color,
hatching, sound and a double coding (color/sound) for the representation of one
value. Two more maps deployed combinations of color/hatching and color/sound
for the display of bi-variate issues. In each map the lowest or largest value had to be
determined. Effectiveness was evaluated via the correctness of the answers, while
efficiency was expressed by the ratio of correct answers and used processing time.
Because goal of the study was to identify basic problems of acoustic encoding
and differences compared to the graphical counterpart, a sample size of 18 test
persons was seen as being large enough.
The key findings of this study can be summarized as follows:
• Not surprisingly, color coding has been proven the most usable method for
extracting quantitative information, while sound coding showed worst results.
On the other hand, a double coding of color and sound was clearly superior
against the hatching option with respect to effectiveness (while efficiency was
slightly worse). An additional value of the double coding approach was seen
through the fact that sound was a good alternative for discriminating between
very similar values which is due to the better resolution of the sound variable
pitch as mentioned before.
• Considering the display of two map topics, the combination of color/hatching
showed a better usability compared to color/sound. There was, however, an
important application case for the color/sound combination: If a couple of small
object or reference areas exist, sonification becomes meaningful because hatch-
ing is not appropriate due to space limitations.
• Observing the heterogeneous and partly not comprehensible interaction patterns
in sound maps, we conclude that a significant portion of the worse usability
effectiveness and efficiency is due to the unfamiliar coding method. Vice versa,
we also observed a significant learning effect throughout the study, despite of an
increasing complexity of tasks. These findings correspond very well with results
